Skip to main content

Full text of "Business Cycles Vol II"

See other formats 


TEXT  FLY  WITHIN 
THE  BOOK  ONLY 


u<  OU_1  60543 


OTJP— 7RR— 13-6-7r>— 5,000 

OSMAN1A  UNIVERSITY  LIBRARY 

Call  No.  '  Accession  No. 

Author 

Title   . 

This  book  should  he  returned  on  »r  hefoic  the   date  last   marked  below. 


BUSINESS  CYCLES 

VOLUME   II 


BUSINESS  CYCLES 

A  Theoretical,  Historical, 

and  Statistical  Analysis  of  the 

Capitalist  Process 


BY 

JOSEPH  A.  SCHUMPETER 

Professor  of  Economics,  Harvard  University 

VOLUME  II 


FIRST  EDITION 


McGRAW-HILL  BOOK  COMPANY,  INC. 

NEW  YORK  AND  LONDON 
1939 


COPYRIGHT,  1939,  BY  THE 
McGRAW-HiLL  BOOK  COMPANY,  INC. 


PRINTED  IN  THE  UNITED  STATES  OP  AMERICA 

All  rights  reserved.  This  book,  or 

parts  thereof,  may  not  be  reproduced 

in  any  form  without  permission  of 

the  publishers. 


THE  MAPLE  PRESS  COMPANY  YORK,  PA. 


Contents 

VOLUME    II 


CHAPTER  VIII 

THE  PRICE  LEVEL 449 

A.  A  Preliminary  Warning  Concerning  the  Causal  and  the  Sympto- 
matic Importance  of  Variations  in  Price  Level 449 

B.  The  Theory  of  the  Price  Level 452 

C.  The  Practical  Question 458 

D.  Analysis  of  the  Behavior  of  Price-level  Series;  the  Pulse  Charts;  the 
Gold  Factor 461 

E.  Group  Prices;  the  Salient  Fact  of  Covariation 475 

CHAPTER  IX 

PHYSICAL  QUANTITIES.     EMPLOYMENT 483 

A.  Individual  and  Composite  Quantities;  Three  Methods  of  Indicating 
Variations  in  Total  Output 483 

B.  The  Analysis  of  the  Trend  in  Total  Industrial  Output;  the  Problem 

of  Retardation 491 

C.  The  Cyclical  Behavior  of  the  Physical  Volume  of  Production   .    .    .  500 

Behavior  of  output  and  price  level  compared 506 

D.  Employment  of  Labor 509 

Discussion  of  the  English  unemployment  series;  absorption  of  cyclical  unemployment  at 

riaing  wage  rates,  cyclical  behavior ...  516 

CHAPTER  X 

PRICES  AND  QUANTITIES  OF  INDIVIDUAL  COMMODITIES 520 

A.  Prices    and    Quantities    of    Individual    Commodities     (Including 
Services) 520 

1.  Expectation  for  infinite  variety  of  responses  to  cyclical  situations 521 

2.  The  competitive  case;  the  "nominal"  effect  of  cyclical  variations  in  expenditure   .    .  522 

3.  Reaction  of  competitive  industries  to  "real"  changes  in  receipts 524 

4.  Illustrative  discussion  of  various  types  of  behavior 525 

B.  Special  Cases 528 

Recent  work  on  price  analysis 528 

The  case  of  coffee 529 

The  hog  cycle  and  other  "animal  cycles" 530 

C.  The  Cycle  in  Shipbuilding;  Professor  Tinbergen's  Model 583 

v 


vi  CONTENTS 

PAGB 

D.  Entrepreneurial  Price  Policies 535 

The  case  of  entrepreneurs  confronted  by  a  demand  of  infinite  elasticity 535 

The  imperfectly  competitive  situation;  oligopoly  in  particular .  536 

Rigidity  or  stability  of  price .                           538 

CHAPTER  XI 

EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES 544 

A.  Some  Propositions  about  Money 544 

B.  System  Expenditure  (Outside  Clearings);  the  Cyclical  Behavior  of 
Producers'  and  Consumers'  Expenditure * 548 

C.  National  Income  and  Wages 561 

1.  The  behavior  of  national  income  as  to  trend  and  cyclical  variations;  comparison  with 
other  series;  the  profit  item ....  561 

2.  Factual  and  conceptual  difficulties  about  wages,  various  definitions 564 

8.  Analysis  of  trends  in,  and  cyclical  variations  of,  wage  bills  and  wage  rates ,  567 

4.  The  question  of  lags ...           .    .                   .  572 

5.  A  difficulty  in  theoretical  interpretation. 574 

D.  Deposits  and  Loans 578 

1.  Possible  ways  of  financing  an  act  of  expenditure.    .                       ...                           .  578 

2.  The  relation  between  (outside)  deposits,  loans,  and  clearings  .  .581 

3.  Qualification  of  the  expectation  for  parallelism .  584 

4.  On  underspending,  nonborrowing,  and  the  role  of  saving  in  the  cycle.  587 

5.  Customers'  balances  and  bank  loans  in  detail;  behavior  of  various  monetary  ratios     .  594 

6.  Real  and  monetary  investment  ...           598 

CHAPTER  XII 

THE  RATE  OF  INTEREST 602 

A.  Earlier  Argument  Resumed 602 

1.  The  concepts  of  a  demand  for  balances  and  of  an  equilibrium  rate  of  interest;  the 
adapted  rate   ...            .    .                         .                     602 

2.  The  secondary  wave  and  the  cyclical  shifts  in  the  demand  for  balances;  the  lag  of 
interest ...                                                          .  604 

8.  The  "supply"  of  balances .    .  606 

4.  The  position  of  interest  in  the  cyclical  process;  capitalisation  of  quasirents  .    .             .  607 

5.  The  pricing  of  titles  to  future  balances 611 

B.  Discussion  of  Various  Rates 614 

The  historical  course  of  interest  rates 614 

1.  Factual  difficulties .           .    .  615 

2.  The  mortgage  rate ...  617 

3.  Bond  yields 618 

4.  The  rates  of  the  open  market 622 

5.  The  rates  of  the  Bank  of  England  and  of  the  Reichsbank 625 

C.  Discussion  of  the  Time  Shape  of  Interest  Rate 626 

1.  The  absence  of  trend  in  the  series 627 

2.  The  cyclical  behavior  of  interest     ...                                                      628 

8.  Interest  and  profits,  pig  iron  production,  total  output,  and  price  level       632 

4.  The  consequential  character  of  interest  and  Professor  von  Hayek's  theory;  the  lag  in 

short  and  long  rates 634 

CHAPTER  XIII 

THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE 639 

A.  Banks  and  the  Pulse  of  Industry 639 

1.  Limitations  to  the  initiative  of  banks;  banks  are  normally  not  "loaned  up"                .  639 

2.  Member  banks'  operations  in  the  open  market;  their  secondary  reserves,  investments, 

and  loans 643 


CONTENTS  vii 

PAGE 

8.  Other  factors  affecting  banks'  investments;  the  investments  of  industrial  or  promotion 

banks 646 

B.  The  Central  Market  (in  an  Isolated  Domain) 648 

Position  and  influence  of  bankers'  banks;  their  behavior  with  respect  to  cyclical  situations  651 

C.  The  Cyclical  Aspects  of  International  Relations 666 

1.  Exports  and  imports  under  the  influence  of  cyclical  fluctuations;  the  role  of  central 
banks 666 

2.  Capital  movements  and  central  bank  policy;  the  English  case;  pivotal  importance  of 
short  claims  on  foreign  countries .               ....  668 

3.  The  London  gold  market;  the  gold  policy  of  the  Bank  of  England;  gold  stock,  bank  rate, 

and  price  level 675 

D.  Stock  Exchange  Series 678 

1.  Cyclical  nature  and  cyclical  effects  of  the  trade  in  stocks  and  bonds;  peculiarities  of  the 
pricing  of  stocks ...                        .                        .                    679 

2.  Theory  of  stock-market  "  tendencies "    ...            .                 .  682 

3.  Analysis  of  the  behavior  of  stock  prices;  various  covariations .  684 

4.  Banks  and  stock  speculation .                   .    .           689 

CHAPTER  XIV 

1919-1929 692 

A.  Postwar  Events  and  Postwar  Problems  (Introduction) 692 

B.  Comments  on  Postwar  Patterns 695 

1.  Some  factors  and  symptoms  of  change  in  the  social  structure 697 

2.  The  world  war  and  its  consequences  as  external  factors         700 

3.  Survey  of  postwar  foreign  policies  and  economic  relations;  reasons  for  the  failure  of  the 
provisional  arrangements  arrived  at           ....  702 

4.  Various  other  types  of  war  effects  exemplified  .    .                      .                .  704 

5.  Postwar  protectionism  and  the  refusal  of  this  country  "to  accept  its  creditor  position"  705 

C.  Further  Comments  on  Postwar  Conditions  in  Our  Three  Countries  708 

1.  The  United  States;  mentality  of  the  country  and  its  fiscal  policy             .                    .  708 

2.  Digression  on  the  effects  of  taxation .  710 

8.  Germany;  the  postwar  situation  and  the  social  atmosphere,  expenditure  and  fiscal  policy  714 

4.  Facts  and  theory  of  the  German  "capital  imports"         .  718 

5.  England;  the  postwar  situation;  monetary  policy;  the  social  atmosphere;  fiscal  policy  722 

D.  Outlines  of  Economic  History  from  1919  to  1929 782 

1.  Agrarian  developments;  diagnosis  of  the  agrarian  depressions  in  the  United  States, 
England,  and  Germany .            .  782 

2.  The  building  booms  of  the  twenties  in  the  United  States,  Germany,  and  England   .  743 

E.  The  "Industrial  Revolution'*  of  the  Twenties 753 

General  features ...  753 

1.  English  developments;  external  factors;  motorcars;  iron  and  steel,  electricity;  the 
expanding  industries;  cycles  and  phases     ...                                                 ....  755 

2.  German  developments,  "rationalization";  public  enterprise,  mergers,  the  "crisis  of 
concerns";  the  potash  problem;  iron  and  steel;  organizational  innovations;  the  chem- 
ical industry;  rayon  in  particular;  cycles  and  phases  .    .  759 

8.  Developments  in  the  United  States;  the  general  pattern;  aviation;  power  production, 
electrical  equipment;  motorcars,  oil  and  rubber,  the  chemical  industry,  rayon  and 
textiles;  iron  and  steel,  aluminum,  copper;  value  added  and  value  added  divided  by 
payroll  in  industries,  whose  value  added  was  50  millions  or  more  in  1929;  detailed 

discussion  of  cyclical  fluctuations  through  1929 785 

F.  The  Behavior  of  Systematic  Series  from  1919  to  1929 794 

I.  Output  (a),  prices  (6),  interest  rates  (c).                ...                    .                    ...  794 

c.  1.  Total  industrial  output  and  output  per  employee          .                          .  795 
2.  Producers'  and  consumers'  goods;  equipment  goods;  "overproduction"  and 

"excess  capacity"            .                          .                                          799 


viii  CONTENTS 

PAGE 

8.  Employment  and  the  absorption  of  unemployment 803 

6.  Price  levels  as  shaped  by  the  evolutionary  process  and  by  reaction  to  war  inflation  807 

1.  Analysis  of  variations  in  the  American  price  level 808 

2.  The  English  case 810 

8.  The  German  case 811 

e.  Interest  rates         811 

1.  The  behavior  of  interest  rates  in  the  United  States 812 

2.  The  German  case 815 

3.  The  English  case  ....                       816 

II.  a.  Debits  outside  New  York,  national  income,  and  payroll  in  the  United  States  .    .  817 

6.  Corporate  accumulation  in  the  United  States  .               820 

e.  Total  realized  income  and  consumers'  expenditure;  households'  borrowings;  infer- 
ence about  households'  savings ...  823 

d.  German  and  British  national  incomes,  wage  bills,  consumers'  expenditures,  and 
savings 827 

e.  The  behavior  of  profits  in  the  United  States 830 

/.  The  behavior  of  wage  rates 834 

1.  American  wage  rates  during  the  twenties 835 

2.  Their  relation  to  the  amount  of  unemployment  and  to  prosperities   ....  838 

3.  German  wage  rates  during  the  twenties 844 

4.  Wage  rates  in  the  United  Kingdom  during  the  twenties       847 

III.  Series  descriptive  of  processes  in  the  sphere  of  money  and  member  bank  credit,  mainly 

in  the  United  States .    .            850 

a.  The  nature  of  time  deposits.    .    .                ...        , 851 

b.  Outside  deposits  and  loans  and  the  structural  change  in  assets 857 

c.  Banks'  investments  and  general  business  practice 862 

d.  German  and  English  banking  developments             .    .  868 

IV.  Stock  speculation  and  the  monetary  investment  process,  mainly  in  the  United  States  870 

a.  The  theory  of  brokers'  loans   ...        .            .        .    .            870 

6.  The  speculative  outburst;  call  rate;  minor  points;  issues  for  real  investment.    .  874 

c.  Stock  exchange  and  investment  processes  in  Germany  and  in  England       .       .    .  880 

V.  Central  banking  in  the  United  States  and  England                .        .  885 

a.  The  American  postwar  situation .            ...            .    .  885 

6.  The  mechanics  of  Federal  reserve  credit .    .           .        .    .  888 

c.  The  reserve  system's  postwar  policy;  conclusions;  its  "failure  to  prevent  the 
depression"         .    .                .                    .                .    .                        ...  894 

d.  Comments  on  the  policy  of  the  Bank  of  England                   ...                .  904 

CHAPTER  XV 

THE  WORLD  CRISIS  AND  AFTER 906 

A.  The  World  Crisis  and  the  Cyclical  Schema 906 

B.  1930 911 

1.  The  United  States                      911 

2.  England       ...                         917 

3.  Germany ...        .  920 

C.  1931  and  1932 924 

1.  Physical  Production,  on  the  downgrade  and  at  the  bottom                   .    .            ....  924 

2.  Incidents,  Accidents,  and  Policy  in  Germany    .        ...           .    .                    930 

3.  Incidents,  Accidents,  and  Policy  in  the  United  States                            936 

4.  The  behavior  of  other  American  and  German  time  series  around  the  bottom  of  the 
depression 944 

D.  The  United  Kingdom,  1931-1938 954 

1.  The  abandonment  of  the  Gold  Standard 955 

2.  The  subsequent  management  of  money  and  credit       958 

8.  Some  effects  of  abandonment  and  cognate  policies  .              959 

a.  The  relief  to  exports 960 

b.  England's  conversion  to  protectionism  and  the  Ottawa  achievement       961 

4.  England's  industrial  processes;  the  great  building  boom  in  particular                    .    .    .  963 

5.  The  testimony  of  time  series,  especially  of  output,  unemployment,  prices,  wages,  in- 
come, clearings,  issues,  stock  prices 966 


CONTENTS  ix 

PAQB 

E.  The  State-Directed  Economy  of  Germany,  1933-1938 971 

Employment  and  production 971 

Efforts  toward  autarky 973 

Policy  in  the  recovery  phase 974 

Government  expenditure  in  the  prosperity  phase         .           .    .               .            ....  976 

Prices,  wages,  incomes,  consumption ...               977 

Monetary  management  and  the  behavior  of  monetary  and  banking  series       979 

F.  Recovery  and  Recovery  Policy  in  the  United  States  from  1933-1935  983 

The  problem .  .983 

1.  "  Minor "  measures  surveyed ...                        .            .  986 

2.  Recovery  and  the  AA A.    .           .    .               .           .    .           .               988 

3.  Recovery  and  the  NRA        .               .                           .  992 

4.  Monetary  policy     ....           ...                       .           .           .  996 

5.  Income-generating  expenditure,  facts  and  theory                    .                                    .    .  1001 

6.  The  statistical  picture ...                .                         .    .  1006 

G.  The  Disappointing  Juglar 1011 

The  problem  raised  by  American  developments  since  1935 ,  .1011 

1.  Time  series  contours           .                .  1013 

2.  The  money  market  and  the  banking  sphere;  behavior  of  price  level  1016 
8.  The  industrial  process .                                         .             .  1020 

4.  The  reciprocal  trade  agreements,  monetary  "restriction"  and  "expansion";  the 
stoppage  of  income  generation                                                      .                         .            .  1026 

5.  Stagnation  of  the  capitalist  process     .               .                                .  1032 

a.  {The  theory  of  vanishing  investment  opportunity                                       .            .    .  1032 

b.  Its  inadequacy       1036 

c.  The  alternative  explanation  by  means  of  the  social  atmosphere  resulting  from  cap- 
italist evolution 1038 

d.  Reasons  for  believing  this  explanation  to  be  adequate 1044 

APPENDIX:   DESCRIPTION   OF  THE  STATISTICAL  MATERIAL  EMBODIED   IN 

CHARTS  I-LX 1051 

Index 1079 


CHAPTER  VIII 

The  Price  Level 


A.  A  Preliminary  Warning. — The  fact  that  price-level  series  are  the 
first  to  be  discussed  should  not  be  interpreted  to  mean  that  we  consider 
them  first  in  either  causal  or  symptomatic  importance.  Businessmen, 
politicians,  and  many  economists  unite  in  drawing  a  picture  which  grossly 
exaggerates  the  role  of  price  movements  in  the  cyclical  process.  While 
for  obvious  reasons  there  is  some  excuse  for  this  in  the  case  of  businessmen 
and  politicians,  only  faulty  analysis  can  account  for  it  in  the  case  of  econo- 
mists. The  very  definition,  "the  crisis  is  a  break  in  prices,"  and  still 
more  propositions  such  as  "the  collapse  of  the  price  system  is  the  real 
cause  of  a  depression,"  betray  failure  to  realize  that  the  cycle  is  a  process 
within  which  all  elements  of  the  economic  system  interact  in  certain  char- 
acteristic ways  and  that  no  one  element  can  be  singled  out  for  the  role  of 
prime  mover.  The  mistake  involved  is,  thus,  much  more  fundamental 
than  it  would  be  if  there  really  were  any  sense  in  searching  the  system  of 
economic  quantities  for  any  single  element  responsible  for  the  cycle,  and  if 
the  theory  alluded  to  merely  seized  upon  a  wrong  one. 

It  should  be  abundantly  clear  from  our  theoretical  discussion  in  the 
third  and  fourth  chapters  and  from  our  historical  discussion  in  the  sixth 
and  seventh,  that  price  movements  are  not  the  all-important  factor  in  the 
business  cycle  that  they  are  sometimes  held  to  be.  We  cannot  too  often 
repeat  that  price  movements  are  not  causal  to  the  prosperity  phases  of  our 
process;  that  prosperity  can,  and  sometimes  actually  does,  start  from  a 
falling  price  level;  that  those  innovations  which  "ignite  *'  prosperity  do  not 
presuppose,  though  they  induce,  an  increase  in  prices  but  are  profitable  at 
the  existing  level;  and  that  innovations  which  are  not  profitable  at  the 
existing  level  are  maladjustments  in  the  same  sense  as  are  all  other 
operations  that  pay  only  because  prices  are  rising.  But  rising  prices, 
being  part  of  the  mechanism — as  far  as  innovations  are  financed  by  credit 
creation  and  under  ideal  conditions  of  full  employment — by  which  factors 
of  production  are  directed  toward  their  new  employments,  do  create 
additional  margins  of  profit.  They  have  a  dislocating  influence  which 
directly  and  also  indirectly,  by  inducing  error  and  condoning  incapacity 
and  misconduct,  accounts  for  much  that  happens  in  the  subsequent 

449 


450  BUSINESS  CYCLES 

periods  of  liquidation  and  helps  to  make  them  abnormal,  i.e.,  to  turn  them 
into  depressions. 

Again,  a  fall  in  prices  is  not  the  same  as  a  fall  in  money  earnings,  which 
in  turn  is  not  the  same  as  a  fall  in  real  earnings.  It  is  necessary,  in  order 
to  get  into  a  sound  frame  of  mind  in  matters  of  the  business  cycle,  to  divest 
oneself  of  the  prejudice  that  a  general  fall  in  prices,  such  as  would  normally 
occur  by  virtue  of  the  working  of  the  cyclical  mechanism,  is  in  itself  a 
catastrophe  or  necessarily  productive  of  catastrophes,"  that  falling  prices 
must  always  spell  misery;  that  they  necessarily  increase  the  burden  of 
debt  in  any  sense  in  which  that  would  be  synonymous  with  causing 
trouble;  that  they  are  incompatible  with  prosperous  business;  or  that  they 
are  simply  an  unmixed  evil  which  is  to  be  prevented  at  any  cost  and  which 
can  be  prevented  without  impairing  the  efficiency  of  the  capitalist 
machine.  But  the  cyclical  fall  in  price  level  that  occurs  in  recession  is  an 
element  in  a  process  of  adjustment  to  the  changes  wrought  by  what  hap- 
pened in  prosperity,  and  that  process  deals  harshly  with  many  people. 
As  rising  prices,  so  do  falling  prices  become  an  intermediate  cause  of 
secondary  phenomena,  and  they  may  also  acquire  a  momentum  of  their 
own  and  then  move,  particularly  during  depression,  in  a  way  that  does 
not  lead  to  adjustment  but  spells  additional  disturbance.1 

Our  analysis,  however,  leads  us  to  believe  that  at  least  the  symptom- 
atic value  of  price  movements  should  be  great.  So  it  is,  of  course,  but 
less  so  than  we  might  think.  For,  as  we  shall  see  presently  and  as  is 
obvious  from  common  experience,  neither  prices  of  individual  commodi- 
ties, influenced  as  they  must  be  by  the  particular  conditions  and  policies 
prevailing  in  individual  industries,  nor  the  whole  world  of  prices,  however 
measured,  can  really  be  expected  to  keep  a  consistent  relation  to  other 
series  representing  industrial  conditions  or  to  the  processes  that  lie  behind 
them.  Most  price  series,  to  be  sure,  display  traces  of  the  cyclical  move- 
ment and  on  the  whole  the  association  is  fairly  satisfactory.  But  we  must 
be  careful  about  prediction  in  any  individual  case  and  refrain  from  draw- 
ing far-reaching  conclusions  about  cycles  for  periods  in  which  price  data 
are  practically  all  we  have. 

Series  of  individual  prices  give  rise  to  many  complicated  questions. 
Those  that  refer  to  finished  products  are  fertile  of  practically  insoluble 
difficulties  about  quality,  local  differences,  and  so  on.  However,  these 
series  have  at  least  an  obvious  meaning.  The  same  may,  with  some 

1  The  distinction  between  the  fall  of  prices  in  recession  and  in  depression  may  facilitate 
agreement,  at  least  as  to  diagnosis  if  not  as  to  policy.  So  many  extra-economic  considera- 
tions and  so  many  valuations  of  the  interests  affected  enter  into  the  latter  that  even  perfect 
agreement  in  economic  argument  would  help  but  little  toward  agreement  about  measures. 
Even  the  purely  economic  argument  cannot  be  fully  presented  in  this  book.  Various 
contributions  to  it,  however,  have  been  and  will  be  offered  at  various  turns  of  our  way. 


THE  PRICE  LEVEL  451 

qualifications,  be  said  of  composites  of  prices  of  different,  but  related, 
commodities,  which  we  will  call  Group  Prices.  Everyone  knows  the 
reasons  that  prompt  us  to  construct  price  indices  of  motorcars  in  general 
or  textiles  in  general.  It  is  also  common  knowledge  that  such  composites 
may  be  very  misleading.  We  shall  not  go  into  the  question  of  principle 
involved,  the  roots  of  which  stretch  far  into  general  theory,  but  will 
confine  ourselves  to  one  remark.  An  index  of  this  sort  may  give  a  picture 
that  is  free  from  many  idiosyncracies  of  the  price  movements  of  the  indi- 
vidual commodities  which  enter  into  it  and  may  be  useful  for  many  purposes. 
It  is,  however,  inadequate  for  our  purpose  because  the  internal  shifts  that 
such  an  index  blots  out  may  be  the  essential  thing.  If  innovation  results 
in  one  of  the  commodities  in  such  a  group  pushing  out  another,  the  varia- 
tions of  their  prices  relative  to  each  other  are  fundamental  to  the  under- 
standing of  the  cycle  or  cycles  during  which  they  occur.  Obviously  a 
composite  consisting  of  the  prices  of  cotton,  wool,  silk,  and  linen  textiles 
from  1780  to  1830  would  be  almost  valueless  in  a  study  of  the  business 
cycles  of  that  period.1 

But  the  general  level  of  prices  raises  an  entirely  different  problem. 
Here  it  is  the  very  meaning  of  the  thing  that  becomes  doubtful.  It  has 
even  been  doubted  whether  there  is  any  meaning  to  it  at  all  or,  less 
radically,  whether  the  general  level  of  prices  measures  a  definite  some- 
thing that  exists  as  such  or  is  merely  a  statistical  figure  measuring,  for 
instance,  the  common  movement  of  individual  prices  or — not  necessarily 
the  same  thing — what  is  common  to  the  movements  of  all  individual 
prices.2  Most  statisticians  and  economists  in  the  long  array  of  authors 
from  Dutot  and  Carli  onward  did  not  bother  about  this  at  all  and  went 
ahead  on  such  common-sense  considerations  as  plausibility  or  absurdity  of 
results  and  convenience  of  calculation,  taking  for  granted  the  economic 
meaning  of  the  procedure.  It  is  hardly  uncharitable  to  say  that,  with 

1  It  should  be  observed,  moreover,  that  indices  of  group  prices  will  occasionally  blur 
the  cyclical  picture  in  other  ways  also.     We  need  only  imagine  a  case  in  which  all  the  con- 
stituent prices  display  strong  cycles  identical  in  everything  but  phase.     There  is  plenty  of 
justification  for  the  toning  down  of  the  fluctuations  that  will  result.     But  the  specifically 
cyclical  aspect  of  the  prices  will  be  partly,  in  the  limiting  case  entirely,  lost.     This  should 
be  borne  in  mind  also  in  the  case  of  the  general  price  level,  which  of  course  will  display  a 
smaller  amplitude  of  fluctuations  if,  as  is  unavoidably  the  case,  constituents  do  not  move 
exactly  in  step.     This  is  as  it  should  be  and  will  not  mislead,  provided  we  confine  ourselves 
to  considering  the  general  price  level  as  a  monetary  parameter  only.     But  if  we  took  it  to 
measure  the  average  amplitude  of  fluctuations  in  individual  prices,  we  should  go  completely 
astray. 

2  Businessmen  and  representatives  of  business  interests,  when  they  speak  of  price  levels, 
mostly  mean  only  the  price  of  their  own  product.     In  a  pamphlet  on  the  desirability  of 
stabilization  of  the  price  level  the  writer  found  that  the  expression  was  used  86  times,  with- 
out meaning  once  what  it  ought  to  mean,  or  in  fact  anything  that  can  be  usefully  designated 
by  the  term. 


452  BUSINESS  CYCLES 

the  outstanding  exceptions  of  .Jevons  and  Edgeworth,  they  evolved  and 
applied  methods  of  measurement  without  knowing  exactly  what  it  was 
they  wanted  to  measure.  Progress  has  been  made  nevertheless,  par- 
ticularly by  systematizing  and  analyzing  criteria  of  choice  between  various 
formulae — a  progress  chiefly  associated  with  the  names  of  Irving  Fisher 
(The  Making  of  Index  Numbers,  1922)  and  L.  von  Bortkiewicz  (see 
Geld,  II.  Die  Messung  des  Geldwertes;  Handworterbuch  der  Staats- 
wissenschaften) .  And  other  lines  of  advance  have  been*  opened  up  by 
Pigou,  Haberler,  Frisch,  Leontief,  and  others.  But  the  particular 
question  that  concerns  us  is  as  yet  so  much  debated,  and  in  most  cases 
so  imperfectly  stated,  that  we  must  try  to  state  and  answer  it1  before  we 
can  discuss  in  the  light  of  our  theoretical  expectations  the  behavior  of 
such  indices  as  we  have,  or  use  them  in  "correcting"  or  "deflating" 
individual  prices  or  other  magnitudes  expressed  in  monetary  units. 

B.  The  Theory  of  the  Price  Level. — Imagine,  for  simplicity's  sake, 
an  isolated  society  without  money,  the  economic  life  of  which  merely 
consists  in  the  current  production  of  consumer's  goods  from  original 
means  of  production,  say,  services  of  labor  and  of  land.  These  consumers' 
goods  are  sold  to  the  very  laborers  and  landlords  who  furnish  the  produc- 
tive services.  Money  and  credit  being  absent2  and  unknown,  equilibrium 
ratios  of  exchange  will  establish  themselves  between  all  pairs  of  economic 
goods.  But  there  will  be  no  absolute  prices.  Now  express  all  these 
ratios  in  a  common  unit,  for  example,  by  putting  the  exchange  value  of  an 
arbitrary  quantity  of  any  arbitrarily  chosen  commodity  equal  to  unity. 
Thereby  the  ratios  are  turned  into  absolute  quantities,  which  we  call 
prices.  If  we  want  to  change  from  this  standard  to  another,  we  simply 
divide  these  prices  by  the  price  of  some  quantity  of  the  new  standard 
commodity  in  the  old  system.  We  do  not  need,  however,  any  commodity 
standard  of  this  sort,  but  can  derive  a  unit  by  putting  any  combina- 
tion of  equilibrium  market  values  equal  to  an  arbitrary  figure,  for  example, 
by  ruling  that  the  sum  total  of  market  value  (prices  times  quantities  sold 
in  a  period  of  account)  of  all  consumers'  goods  be  equal  to  100  units  or 
100  billions  of  them.  This  will  uniquely  define  the  meaning  of  this — 
otherwise  meaningless — unit  in  terms  of  every  commodity,  just  as  well  as 
the  choice  of  a  standard  commodity  would.  The  obvious  practical 

1  The  solution  to  be  presented  derives  from  Walras.     It  has  already  been  given  by 
Professor  Francois  Divisia  (L'Indice  Monetaire,  Revue  d'£conomie  Politique,  1925,  and 
Economique  Rationelle,  1928,  pp.  252-280),  to  whom  belongs  entirely  whatever  merit  there 
is  in  the  idea.     Reference  should  also  be  made  to  the  recent  discussion  in  the  Review  of 
Economic  Studies,  vol.  III. 

2  Unless,  indeed,  money  is  defined  by  the  criterion  of  indirect  exchange,  in  which  case 
many  commodities  would  function  as  money. 


THE  PRICE  LEVEL  453 

difficulties  are  not  relevant  to  this  argument.  But,  however  we  proceed 
in  order  to  acquire  the  immense  advantage  of  a  unit  of  calculation  and 
clearing  (a  "unit  of  account")*  we  must  always  introduce  something  that 
is  arbitrary,  both  in  the  sense  that  the  system  of  quantities  of  commodities 
and  exchange  ratios  does  not  itself  determine  it,  and  in  the  sense  that  the 
particular  decision  is — as  long  as  the  same  equilibrium  persists — entirely 
immaterial.  Also,  any  change  would  be  immaterial  if  all  prices  and  all  the 
other  monetary  magnitudes  could  be  instantaneously  and  perfectly 
adapted  to  it. 

The  social  decision  which,  in  order  to  bring  prices  into  existence,  it  is 
necessary  to  add  to  the  other  conditions  that  determine  the  system  of 
economic  quantities  need  not,  of  course,  consist  in  any  conscious  act, 
which  it  would  be  practically  impossible  to  perform.  It  may,  and 
historically  did,  come  about  by  way  of  the  growth  of  a  social  habit,  which 
evolved  the  special — and  logically  rather  abnormal — case  of  the  com- 
modity standard.  What  matters  here  is  merely  the  fact  that  such  an 
arbitrary  choice,  however  it  comes  about  and  whatever  particular  form  it 
takes,  supplies  the  additional  equation  that  we  need  in  order  to  have 
uniquely  determined  absolute  prices.  Without  a  unit's  being  given  in 
which  to  express  prices  or,  as  we  may  now  say,  without  the  choice  of  a 
level  of  prices,  these  would  be  indeterminate  because  it  is  only  their  rela- 
tions to  each  other  which  are  determined  by  the  system  itself. 

Hence  the  price  level,  or  monetary  parameter,  is  not  a  mere  statistical 
aggregate  or  a  mean  like  the  average  height  of  recruits  of  a  given  age  in  a 
given  population,  but  a  real  thing  existing  independently  of  the  statistician 
and  to  be  distinguished  from  the  relations  of  prices  to  each  other,  which 
we  shall  designate  as  the  price  system.1  It  is  not  of  the  same  nature  as  a 
group  index,  and  is  more  than  merely  the  most  comprehensive  of  them. 
Considerations  of  convenience  of  calculation  enter  no  doubt  into  the 
practical  method  of  its  measurement  in  a  secondary  way,  but  not  into  its 
concept.  There  is,  in  the  same  sense,  no  question  of  weighting  or  averag- 
ing. Considerations  of  plausibility  have  no  place  at  all.  The  various 
tests  of  index  numbers  have  to  be  replaced,  for  our  purposes,  by  the  single 
question  whether,  and  how  accurately,  a  given  formula  expresses  the 
changes  in  that  parameter.  It  follows,  moreover,  that  the  price  level, 
so  defined,  is  not  itself  a  price  and  cannot  usefully  be  described  in  terms  of 
supply  and  demand,  for  these  categories  apply  only  within  the  world  of 
individual  commodities.  And  it  also  follows  that  considerations  of 
utility  or  welfare  or  of  the  so-called  subjective  value  of  money,  which  may 
move  in  opposite  directions  for  people  with  different  budget  combinations, 

1  This  term  is  used  in  the  same  sense,  we  believe,  by  Professor  Mills  in  Behavior  of 
Prices. 


454  BUSINESS  CYCLES 

are  not  relevant  either  to  our  concept  of  price  level  or  to  the  method  of 
measuring  it.1 

That  social  decision  then  fixes  the  price  level  for  the  equilibrium 
state  of  the  economic  system  which  obtains  at  the  moment;  but  it  would 
stay  as  thus  fixed  only  in  a  perfectly  stationary  society.     In  that  case, 
however,  there  would  be  as  little  meaning  to  the  question  what  the  value 
of  the  price  level  is  as  there  would  be  to  the  analogous  question  in  the  case 
of  a  potential.     In  reality  the  price  level  changes  all  the  tim«  without  any 
change  in  the  social  decision  itself.     For  practically  every  change  that 
occurs  in  the  economic  process  affects  it,  and  it  is  extremely  unlikely  that 
changes  will  occur  precisely  in  such  a  way  as  to  compensate  their  effects  on 
it.     In  particular,  if  nothing  at  all  happens  in  the  sphere  of  money  and 
credit,  the  price  level  will  nevertheless  undergo  variations.     And  there  is 
not  only  meaning  but  obvious  importance  to  the  problem  of  measuring 
them.     This  problem  would  be  easy  to  solve,  if  the  price  system  did  not 
change,  i.e.,  if  prices  never  changed  except  proportionally.     We  could 
then  read  off  the  change  in  the  price  level  from  the  change  in  any  one 
price.     Unfortunately,   however,   the  price   system   also   could   remain 
constant  only  in  a  stationary  society.     As  a  matter  of  fact,  it  changes  in 
time  just  as  frequently  as  the  price  level.     Hence,  the  price  of  any  indi- 
vidual commodity,  as  we  observe  it  at  any  point  of  time,  must  be  inter- 
preted as  the  result  of  two  distinct  components:  the  price  level  and  the 
price  system.     It  is  also  easy  to  see  that  changes  in  the  level  can  in  prac- 
tice hardly  ever  come  about  except  by  way  of  changes  in  the  system — 
even  as  changes  in  the  system  in  practice  hardly  ever  come  about  without 
enforcing  a  change  in  the  price  level.     Yet  the  two  components   of 
change,  however  inextricably  mixed,  are  logically  distinct.     We  have  not 
understood  a  given  change  in  prices  in  general  or  in  any  single  price  as 
long  as  we  have  not  quantitatively  separated  them.     Hence  the  question: 
Seeing  that  we  have  only  the  actual  prices,  is  there  any  means  of  doing  so  ? 
If  the  system  changed  as  well  as  the  level,  while  the  quantities  of  all 
commodities  remained  constant,  the  variations  of  total  actual  expenditure 
would  exactly  measure  the  changes  in  the  price  level,  whatever  happened 
to  the  individual  prices.     If  quantities  also  change,  the  problem  becomes, 
strictly  speaking,  insoluble.     But  here  the  differential  method  of  analysis 
comes  to  our  rescue.     Given  the  usual  conditions  of  differentiability  in  the 

1  This  is  important  to  bear  in  mind,  because  recent  discussion  has  precisely  turned  on 
the  utility  aspect  in  order  to  define  equivalence  of  different  collections  of  income  goods  in 
different  points  of  time.  This  equivalence,  in  turn,  is  made  to  yield  a  criterion  for  the 
change  that  may  have  occurred  in  the  "purchasing  power"  of  the  monetary  unit.  From 
this  standpoint,  Professor  Haberler  has  very  naturally  been  led  to  deny  the  existence  of  any 
general  price  level.  But  all  this,  important  as  it  is  for  other  purposes,  does  not  concern  us 
here,  and  argument  on  these  lines  is  as  irrelevant  for  our  purpose  as  our  argument  is  for 
welfare  considerations. 


THE  PRICE  LEVEL  455 

(smoothed)  time  sequence  of  the  relevant  magnitudes,  we  may  still 
disentangle  the  rate  of  change  of  the  price  level  at  any  given  moment  of 
time.  For  this  purpose,  we  start  from  the  expenditure  on  every  one 
of  our,  say,  n  commodities,  equal  to  its  price  p*  times  its  quantity  q* 
(i  =  1  •  •  •  ri)  bought  at  any  time,  form  the  total  differential  (rate  of 
change  of  expenditure) 


and  sum  over  all  commodities,  so  as  to  get,  if  total  expenditure  be  E, 

{  —  n  i**n 

dE  =        q4pi  +  2  P^dq^ 


Now,  the  dq's  being  the  increments,  positive  or  negative,  of  the  quantities 
of  commodities,  ^Lp^dq^  that  is,  these  increments  times  the  "old"  prices, 
gives  approximately  that  part  of  the  variation  in  expenditure  which  is 
balanced  as  to  its  effects  on  the  level  by  the  quantity  changes,  and  by 
which  expenditure  would  have  had  to  change  in  order  to  keep  prices 
constant,  if  their  system  had  not  changed,  or  at  least  the  price  level,  if  the 
system  has  changed.  To  put  it  in  still  another  way,  if  expenditure  had 
changed  exactly  by  ^ptdqi9  so  that  2qtdpi  =  0,  that  part  of  the  new  total 
of  expenditure  would  have  remained  constant  which  does  not  buy  the  new 
positive  or  negative  increments  and  may  be  looked  upon  as  expended  on 
the  same  kinds  and  quantities  of  commodities  as  before.  But,  with 
unchanging  kinds  and  quantities  of  commodities,  constant  expenditure 
defines  identical  price  levels.  And  as  there  cannot  be,  at  one  time,  two 
price  levels  with  respect  to  the  same  commodities,  the  price  level  would  be 
the  same  as  before.  Hence,  for  a  change  in  the  price  level  to  occur,  it  is 
necessary  and  sufficient  that  Sq+dpi  ^  0,  and  the  departure  from  zero  of 
this  quantity,  therefore,  approximately  measures,  by  its  relation  to 
2pt#»,  the  change  that  actually  occurred. 

We  see  the  principle.  It  consists  in  reducing  the  unmanageable, 
but  general,  case  —  that  of  simultaneous  change  in  level,  system,  and 
quantities  —  to  the  manageable  case  of  unchanging  quantities,  which  is 
always  implied  in  the  general  case  and  can  be  extricated  from  the  rest 
if  changes  are  small.  This  means,  of  course,  that  the  solution  is  but  an 
approximation  and  that  any  method  based  upon  it  breaks  down  if  the 
changes  in  individual  prices  or  quantities  cannot  be  decomposed  into  small 
ones.1  It  also  means  that  comparisons  of  levels  are  possible  only  between 

1  It  breaks  down,  therefore,  in  times  of  sudden  and  violent  changes,  such  as  extreme 
inflation.  It  does  not,  however,  break  down  because  of  the  introduction  of  new  commodi- 
ties, provided  they  do  not  suddenly  intrude  in  quantities  that  are  big  with  reference  to  all 
the  others.  How  far  an  index  of  this  nature  fulfills  Professor  Fisher's  tests  cannot  be  dis- 


456  BUSINESS  CYCLES 

neighboring  points  of  time,  and  that  the  states  of  things  obtaining  at 
finitely  distant  points  of  time  cannot  be  compared  directly  but  only  by 
way  of  all  intermediate  points.  In  practice,  of  course,  smallness  of  dis- 
tance may  be  interpreted  somewhat  less  strictly.  Dividing  monetary 
magnitudes  by  an  index  of  this  kind  means  the  elimination  from  them  of 
the  influence  of  changes  in  the  significance  of  the  monetary  unit  i.e., 
of  a  factor,  the  action  of  which  makes  their  comparison  all  but  meaningless 
and  which  is  itself  meaningless  insofar  as  the  really  relevant  elements  of 
economic  life,  quantities  of  commodities  and  ratios  of  exchange,  are  con- 
cerned. This  is  what  is  here  meant  by  "deflating**  sequences  of  items 
expressed  in  "current  dollars.'*1 

Next  we  have  to  find  out  which  of  the  better  known  formulae  may 
be  considered  satisfactory,  or  approximately  so,  from  the  standpoint 
of  this  theory.  Of  course  we  are  not  concerned  with  any  of  the  many 
special-purpose  indices,  which,  as  the  reader  will  realize  by  now,  have 
nothing  to  do  with  measuring  the  price  level  in  our  sense,  although  the 
word  level  invariably  occurs  in  association  with  them,  and  although  their 
construction  may  be  amply  justified  by  the  purposes  they  are  to  serve. 

Happily  we  make  a  very  comforting  discovery.  One  of  the  most 
common  formulae  —  common,  it  is  true,  more  in  the  theory  of  the  subject 
than  in  practical  index  making  —  is  the  one  usually  referred  to  as  the 
Laspeyres  formula.  It  compares  two  aggregates:  the  quantities  of  the 
base  period  times  the  prices  of  the  current  period  and  the  same  quantities 
times  the  prices  of  the  base  period.  If  we  denote  any  price  or  quantity 
of  the  current  period  by  the  subscript  i,  and  any  price  or  quantity  of  the 
base  period  by  the  subscript  o,  the  formula  is  2ptq0/2p0qo-  It  is  obvious, 
and  has  often  been  pointed  out,  that  this  formula  imparts  an  increasing 
bias  the  farther  away  we  draw  from  the  base  period.  But  if  we  have,  for 
example,  monthly  data,  which  will  in  most  cases  represent  "small" 
intervals,  and  if  we  refer  each  item  to  its  predecessor  as  its  base  ("chain- 
method,"  first  effectively  advocated  by  A.  Marshall),  then  we  may  put 
each  pi  —  p0  +  dp0  and  get,  dropping  subscripts  : 

2(p  +  dp)q  _  1       Zqdp 


cussed  here.     Nor  is  this  question  relevant  for  our  purpose.     But  it  may  be  stated  that  it 
fulfills  all  the  criteria  that  retain  meaning  within  this  theory. 

1  Objections  to  the  performance  of  that  operation  have  been  raised  by  many  eminent 
authors.  They  are  but  too  well  founded  if  they  rest  on  the  quality  of  available  indices. 
They  are  also  well  founded  if  they  aim  at  erroneous  conclusions  arising  from  an  imperfect 
understanding  of  what  "deflating"  really  means.  But  the  operation  remains  nevertheless, 
as  unavoidable  as  it  is  theoretically  correct  if  properly  understood.  It  has  been  pointed 
out  to  the  writer,  and  it  may  well  be  true,  that  the  use  of  the  term  price  level  for  our  concept 
may  prove  misleading.  Mr.  R.  Bryce  has  suggested  the  term  monetary  parameter. 


THE  PRICE  LEVEL  457 

This  is  our  own  formula.  We  could  also  have  followed  the  suggestion  of 
Paasche  and  compared  the  aggregate  of  expenditure  at  the  current  period 
with  the  sum  which  would  have  been  expended  at  the  base  period  if  at  the 
then  prevailing  prices  (p0)  the  current  quantities  (gt)  had  been  bought. 
Hence  the  suggestion  of  Knut  Wicksell  (which  he,  however,  based  only  on 
the  absence  of  a  criterion  of  choice  between  the  Laspeyres  and  the  Paasche 
formula),  namely,  to  take  the  geometric  mean  between  the  two  (Professor 
Fisher's  Ideal  Formula)  is,  although  not  quite  in  accordance  with  our 
theory,  also  acceptable  provided,  of  course,  that  changes  are  small. 

Finally,  there  is  the  question  which  prices  to  include.  If  the  price  level 
in  our  sense  is  a  definite  thing,  the  answer  must  follow  from  the  theory  of 
that  thing.  So  it  does.  We  ought  to  include  the  prices  times  quantities 
of  all  commodities  and  services  actually  and  directly  bought  by  house- 
holds, and  nothing  else.  A  very  simple  argument  will  establish  this 
principle.1 

The  price  level  in  our  sense  is  a  measure  of  a  property  of  the  system  of 
economic  values.  This  parameter  derives  from  the  relation  between  the 
flow  of  expenditure  and  the  flow  of  the  things  bought  by  it,  and  thereby 
defines,  in  a  particular  way  and  for  a  particular  purpose,  the  significance 
of  the  unit  of  accounting  and  clearing  in  terms  of  commodities  and  services. 
Now  this  flow  of  expenditure  runs,  as  it  were,  through  several  basins 
or  economic  spheres.  For  our  present  purpose  we  may  reduce  these  basins 
or  spheres  to  four:  the  "markets"  of  finished  consumers'  goods,  of  original 
means  of  production  (primarily  labor),  of  produced  means  of  production 
(primarily  raw  materials  and  machinery),  and  of  titles  to  income  (pri- 
marily shares,  bonds,  and  realty) .  Expenditure  flows  through  all  of  these 
but  each  element  of  it  confronts,  at  any  one  time,  not  all  but  only  one  of 
them.  And  there  is  no  meaning  to  a  combination  of  items  from  different 
spheres,  or  the  whole  of  all  the  items  of  different  spheres  or  phases,  of  the 
monetary  stream.  A  variation  in  one  direction  of  a  price  in  the  market  of 
consumers'  goods  is  not  compensated  by  an  equivalent  variation  in  the 
other  direction  of  a  price  in  the  market  of  producers'  goods.  There  is,  to 
be  sure,  plenty  of  interdependence  between  the  different  spheres,  and  units 
of  potential  expenditure  can  be  shifted  from  one  to  the  other.  But  this  is 
irrelevant  for  the  arithmetic  of  the  thing.  The  relevant  criterion  is  sub- 
stitutability  in  the  technical  sense;  we  must  combine,  in  order  to  get  the 
proper  price-level  figure,  the  prices  and  quantities  of  all  goods  which 
compete  for  the  sum  that  actually  buys  in  a  given  sphere  and  in  a  suitable 
time  interval,  and  nothing  else.  If  we  include  less,  we  may  get  a  change 

1  A  fuller  discussion  will  be  found  in  the  writer's  book  on  money.  It  is  only  the 
principle  that  interests  us  at  the  moment.  In  practical  work,  as  we  cannot  hope  to  include 
everything,  the  problem  becomes  one  of  sampling,  and  considerations  of  probability  enter 
at  this  stage,  although  they  did  not  enter  into  the  theory. 


458  BUSINESS  CYCLES 

which  is  due  to  a  shift  of  values  between  rival  commodities.  If  we  include 
more,  we  vitiate  our  result  by  telescoping  different  stages  or  phases  of  the 
monetary  process  into  one.  This  is  what  sometimes  lies  at  the  bottom  of 
the  argument,  which  has,  however,  another  surface  meaning,  that  we  must 
not  include  wages,  except,  of  course,  the  wages  paid  for  services  directly 
consumed,  because  this  would  involve  "double  counting."  If  we  include 
"everything  purchasable,"  we  get  a  meaningless  heap. 

But  the  only  market  which  exhausts,  and  does  not  more  .than  exhaust, 
a  complete  stage  or  phase  of  the  monetary  process  and  presents  the  liga- 
men  between  money  stream  and  commodity  stream,  is  the  market  of 
consumers'  goods.  The  things  that  form  the  markets  of  "original"  and 
of  "produced"  means  of  production  are  substitutable  for  each  other  at 
many  stages  of  the  productive  and  commercial  process,  and  no  combina- 
tion of  their  prices,  therefore,  ever  displays  the  changes  of  our  parameter 
alone.  Nor  can  we  take  all  producers'  goods  together,  because  they  do 
not  fulfill  the  condition  of  facing  one,  and  only  one,  stage  of  the  stream 
of  expenditure,  but  are  obviously  arrayed  in  successive  groups. 

In  the  case  of  the  sphere  of  titles  to  income,  roughly,  the  stock  exchange 
and  the  realty  market,  there  is  a  special  reason  for  exclusion.  They  do  not 
form  a  phase  of  the  fundamental  stream  of  expenditure  at  all  and,  as  we 
shall  see  later  on,  their  pricing  process  is  so  different  from  the  pricing 
processes  in  the  world  of  commodities  and  services  that,  even  for  the  usual 
aims  of  investigations  into  the  "value  of  money"  and  from  the  standpoint 
of  the  usual  theory  about  the  meaning  of  price  indices,  inclusion  with  any 
significant  weight  would  yield  a  result  incapable  of  serving  any  useful 
purpose. 

C.  The  Practical  Question. — It  would  not  be  impossible  to  build  an 
index  of  the  price  level  in  our  sense  (it  could  obviously  be  an  index  only) 
from  postwar  material.  Even  for  the  prewar  time  information  is  avail- 
able (stretching  in  some  cases  over  centuries1)  that  would  go  far  towards 

1  Stray  data  about  prices  we  have,  of  course,  for  practically  all  periods  of  recorded 
history.  Data  about  individual  prices,  previous  to  1780,  are  sometimes  available  in  long, 
continuous  series;  compare,  for  example,  those  in  Thorold  Rogers'  or  d'Avenel's  works  and 
Sir  William  Beveridge's  annual  series  of  Wheat  Prices  in  Western  and  Central  Europe, 
1500-1869  (see  Weather  and  Harvest  Cycles  in  the  Economic  Journal,  December  1921). 
See  also  Professor  Usher's  article  Prices  of  Wheat  and  Commodity  Price  Indices  for  Eng- 
land, 1259-1930,  Review  of  Economic  Statistics,  August  1931.  For  this  country  several 
pieces  of  work  have  been  published,  notably,  Bezanson,  Gray,  and  Hussey,  Prices  in 
Colonial  Pennsylvania,  1935,  G.  R.  Taylor,  Wholesale  Commodity  Prices  at  Charleston 
1732-1791,  Journal  of  Economic  and  Business  History  vol.  IV;  T.  S.  Berry,  Wholesale 
Commodity  Prices  in  the  Ohio  Valley,  Review  of  Economic  Statistics  for  August  1935;  H.  M. 
Stoker,  Wholesale  Prices  at  New  York  City,  1720-1800,  Memorandum  142,  Agricultural 
Experiment  Station,  Cornell  University;  and  R.  Crandall,  Wholesale  Commodity  Prices 
in  Boston  during  the  Eighteenth  Century,  Review  of  Economic  Statistics,  June  1934.  Back 


THE  PRICE  LEVEL  459 

giving  us  an  idea,  although  too  crude  for  microscopic  purposes,  of  the 
behavior  of  the  level.  This  task,  however,  is  quite  outside  the  range  of 
the  possibilities  of  the  individual  worker. 

Under  these  circumstances,  we  are  at  present  forced  to  use,  and  to 
perform  certain  operations  on,  existing  price  indices,  which  for  a  variety 
of  reasons  bear  but  a  distant  relation  to  what  we  wish  to  study.  If 
we  want  to  cover  at  least  the  period  from  the  eighties  of  the  eighteenth 
century  onward,  we  must  sometimes  use  (although  this  has  been  avoided 
wherever  possible)  annual  and  quarterly  indices,  which  cannot  give  correct 
contours  for  the  shortest  of  our  cycles  or  the  correct  boundaries  or  extrema 
of  the  longer  ones,  besides  violating  our  condition  that  the  intervals  must 
be  small.  Lack  of  necessary  data  about  quantities  is  by  itself  sufficient 
to  reduce  the  value  of  some  indices,  at  some  times  of  all  available  indices, 
to  a  mere  indication  of  the  strongest  features.  In  some  cases  the  data 
themselves  and  the  degrees  to  which  the  price  quotations  used  are  repre- 
sentative of  the  prices  that  really  ruled  (but  often  differed  widely)  in  the 
countries  and  at  the  times  to  which  they  refer,  are  open  to  serious  doubt. 
Changes  in  qualities  very  often  present  insuperable  difficulties.  In 
other  cases  technique  is  deficient  beyond  remedy,  from  whatever  stand- 
point it  may  be  looked  at.1  But  the  old  argument  of  practical  workers 
that  indices  tend  to  give  roughly  the  same  picture,  however  well  or  faultily 
constructed,  contains  after  all  some  little  element  of  truth,  which  for  us, 

to  1680  goes  W.  B.  Weedon,  Economic  and  Social  History  of  New  England,  1620-1789, 
1891.  Of  the  work  done  for  other  countries,  Professor  E.  J.  Hamilton's  on  Spanish  and  on 
French  prices  is  the  most  important.  See  also  A.  P.  Usher's  series  (unfortunately  not 
published  in  full)  of  French  Wheat  Prices,  1350-1788  (The  General  Course  of  Wheat  Prices 
in  France,  1350-1788,  Review  of  Economic  Statistics  for  November  1930),  and  E.  B.  Schum- 
peter,  English  Prices  and  Government  Finance,  Review  of  Economic  Statistics  for  February 
1938.  A  long  list  of  contributions  could  be  presented.  For  the  interpretation  of  such 
series  a  detailed  history  of  the  currency  is,  however,  indispensable. 

1  On  questions  of  technique  and  on  reliability,  as  analyzed  from  the  standpoint  of 
statistical  criteria,  the  reader  should  consult,  from  the  great  literature  of  the  subject,  at 
least  the  standard  works  of  Professor  Irving  Fischer,  The  Making  of  Index  Numbers, 
and  Professor  Wesley  C.  Mitchell,  Index  Numbers  of  Wholesale  Prices  in  the  United  States 
and  Foreign  Countries,  Bulletin  of  the  U.  S.  Bureau  of  Labor  Statistics  173,  and  The  Making 
and  Using  of  Index  Numbers,  Bulletin  of  the  U.  S.  Bureau  of  Labor  Statistics  284;  also 
Professor  Warren  Person's  The  Construction  of  Index  Numbers,  Professor  Mills'  remarks 
on  the  subject  in  The  Behavior  of  Prices  (see,  in  particular,  the  instructive  graphs  on  p.  237 
and  the  section  on  reliability)  and  the  excellent  chapter  on  Index  Numbers  by  the  late 
A.  A.  Young  in  Rietz's  Handbook  of  Mathematical  Statistics.  Any  more  comprehensive 
study  would  have  to  start  from  Edge  worth's  famous  reports,  reprinted  in  his  Collected 
Papers.  For  England,  the  Board  of  Trade  Index;  for  the  United  States,  the  Bureau  of 
Labor  Index  must,  as  far  as  the  writer  has  been  able  to  judge,  be  considered  as  the  best  in 
the  usual  sense,  as  well  as  in  the  sense  that  they  approximate  our  price  level  more  than  any 
others.  On  the  latter,  see  H.  B.  Arthur,  The  Development  of  Wholesale  Price  Measure- 
ment by  the  Federal  Government,  Review  of  Economic  Statistics  for  August  1935. 


460  BUSINESS  CYCLES 

it  is  believed,  suffices  to  justify  what  we  are  going  to  do  with  them, 
provided  we  watch  our  step  in  drawing  conclusions. 

In  one  respect  there  seems  to  be  more  reason  for  confidence  than 
most  students  of  price-index  numbers  feel,  namely,  in  respect  to  the  use 
of  indices  of  prices  at  wholesale  in  lieu  of  the  retail  prices  for  which  our 
own  theory  really  calls.  Owing  to  the  inadequacy  of  the  retail  price 
material  before  the  war,  we  have  no  choice  but  to  use  wholesale  price 
indices,  although  what  is  available  of  the  former  has  been  studied.1  But 
there  are  disadvantages  inherent  in  the  use  of  retail  price  indices  which  no 
critical  care  or  perfection  of  technique  that  can  reasonably  be  hoped  for  in 
the  near  future  can  be  expected  to  overcome.  It  is  certain  retail  prices 
that  are  particularly  "sticky"  in  the  short  run.  Some  are  very  "tradi- 
tional," and  it  is  here  more  than  anywhere  else  that  change  in  quality 
(also  quantity)  takes  the  place  of  change  in  price.  Moreover,  it  is  diffi- 
cult to  discover  and  appraise  changes  in  the  services  rendered  by  retailers 
to  customers  and  the  nature  of  that  composite  of  ponderable  and  impon- 
derable elements  which  is  what  the  household  really  buys,  or  to  follow  up 
the  sometimes  tortuous  ways  of  supernormal,  normal,  and  "sale"  prices. 
Not  always,  but  often,  retail  trade  acts  as  a  bottle-neck  that  prevents  the 
stream  of  commodities  from  flowing  along  its  course.  All  this,  however 
essential  for  other  purposes,  is  distortion  of  the  true  contour  lines  from 
our  standpoint.  Prices  of  commodities  at  wholesale,  on  the  other  hand, 
may  overdraw  the  picture,  but  they  at  least  draw  it.  There  is  some  truth 
in  Mr.  Snyder's  saying  that  they  are  a  picture  of  "speculation"  rather 
than  commercial  reality,  and  they  are  certainly  likely  to  display  peaks  and 
troughs,  which  do  not  mean  much  by  themselves  either  as  to  amplitude  or 
as  to  exact  location.  But  that  particular  kind  of  "  speculation  "  reflects 
the  opinion  of  wholesale  trade  about  imminent  reality  in  the  short  run, 
and  the  real  tendency  of  things  in  the  long  run,  free  from  many  frictions, 
rigidities,  and  inertiae.  Here,  business  life  does  for  us  a  piece  of  work  of 
abstraction  and  analysis  that  could — and  for  purposes  requiring  a  higher 

1  We  are  rapidly  acquiring  valuable  series,  especially  of  the  cost-of -living  type  (which, 
to  be  sure,  is  not  quite  what  we  want  when  investigating  the  changes  of  the  price  level  in 
our  sense).  Most  of  these  are  in  common  use,  but  there  are  many  others,  in  particular  in 
some  countries  not  included  in  this  study.  Far  above  everything  else,  both  as  to  abundance 
and  reliability  of  material  and  as  to  excellence  of  workmanship,  is  Professor  Gunnar 
Myrdal's  Cost  of  Living  in  Sweden,  1880-1930  (Stockholm  Economic  Studies,  1933). 
See  especially  the  chart  on  p.  141,  which  covers  the  period  from  1830  to  1913.  The  Canad- 
ian Index  (R.  H.  Coats)  1900-1915,  some  of  the  material  in  the  English  Report  of  an 
Enquiry  by  the  Board  of  Trade  into  Working-class  Rents  and  Retail  Prices  •  •  •  in  1912 
(Ed.  5,955,  1913),  which  among  other  things  gives  food-price  indices  1900  to  1912  for 
fourteen  countries,  and,  for  this  country,  Mr.  Carl  Snyder's  indices  and  their  constituents 
have  also  been  very  helpful. 


THE  PRICE  LEVEL  461 

degree  of  accuracy  will  have  to — be  done  on  retail  series  only  by  laborious 
and  unsafe  methods. 

Inasmuch  as  retail  prices  are  prices  of  finished  products,  their  relations 
to  the  associated  wholesale  prices  will  differ  according  to  the  conditions 
and  policies  of  the  industry  or  even  of  the  individual  manufacturing 
(sometimes  wholesale)  concerns.  We  find  extremely  variable  patterns 
very  refractory  to  generalization  and  ranging  from  nearly  instantaneous 
(even  rigid — where  wholesale  price  is  a  fixed  percentage  of  retail  price) 
covariation  between  retail  and  wholesale  prices  to  what  in  the  short  run 
almost  amounts  to  independence.  In  the  case  of  foods  (if  we  exclude 
highly  processed  and  branded  types)  we  presumably  come  nearer  than  in 
any  other  case  to  what  seems  to  be  the  (theoretically)  normal  state  of 
things — i.e.,  that  the  change  of  retail  price  should  lag  behind,  and  roughly 
be  equal  in  amount  to,  the  change  (per  corresponding  unit)  in  wholesale 
price,  either  of  the  object  itself  or  some  raw  material  or  semifinished 
ingredient  of  it,  a  rule  that,  of  course,  has  to  be  qualified  by  taking  account 
of  stocks  and  expectations.  Obviously,  this  means  a  smaller  percentage 
change  in  the  retail  price.1  Actually,  covariation  between  indices  of  retail 
and  wholesale  prices  is  more  in  evidence  than  we  have  a  right  to  expect. 
The  reader  can  easily  satisfy  himself  that  cyclical  variations  displayed 
respectively  by  wholesale  and  retail  indices  do  not  differ  sufficiently  to 
cause  serious  concern  about  any  of  the  statements  to  be  illustrated  or 
proved  by  the  former  (see,  for  example,  Mr.  Carl  Snyder's  charts  for  this 
country  and  England  on  pages  390  and  397,  American  Economic  Review 
for  September  1934). 

D.  Analysis  of  the  Behavior  of  Price-level  Series.2 — This  analysis 
starts  by  recognizing  that  they  are — to  use  the  terminology  introduced  in 
the  first  and  fifth  chapters — synthetic,  systematic,  primary,  conse- 
quential, and  that  they  display  a  result  trend.  We  have  first  to  form 
expectations  as  to  their  behavior  from  the  "pure"  model,  as  modified  by 
the  qualifications  which  constitute  the  further  approximations,  then  to 

1  The  above  agrees  with  Professor  Bowley's  results  in  Wholesale  and  Retail  Prices  of 
Food,  Economic  Journal  for  December  1913,  which  to  the  writer  seems  still  to  be  the  leading 
contribution  to  the  subject.  Also  compare  Frances  Wood,  Construction  of  Index  Numbers 
to  show  Changes  in  the  Price  of  the  Principal  Articles  of  Food  for  the  Working  Classes, 
ibid.,  to  which  the  writer  feels  much  indebted. 

a  The  first  author  to  discuss  price  level  in  the  light  of  scientific  principle  was  probably 
W.  St.  Jevons.  But  Tooke  and  Newmarch's  history  of  prices  is  also  a  discussion  of  the 
factors  relevant  to  variations  of  the  price  level.  Although  marred  in  many  particulars  by 
inadequacy  of  theoretical  equipment  and  a  certain  "wooliness,"  that  work  is  still  a  mine, 
not  only  of  material  but  also  of  wisdom,  and  must  be  ranked  very  high.  Analysis  in  the 
light  of  industrial  fact  is  also  the  great  merit  of  W.  Layton's  Introduction  to  the  Study  of 
Prices  (1st  ed.,  1912;  2d  ed.,  1935). 


462  BUSINESS  CYCLES 

look  at  our  material  and  compare,  and  finally  to  see  whether  discrepancies 
can  be  satisfactorily  accounted  for  by  external  factors1  or  by  defects  of  our 
material.  Since  these  factors  and  defects  are  obviously  important  at  all 
times  and  dominant  at  some,  and  since,  moreover,  our  process  is  "inter- 
nally irregular,"  it  would  be  quite  unreasonable  to  formulate  our  task  in 
any  other  way.  We  must  from  the  outset  be  prepared  for  considerable 
discrepancies,  and  all  we  can  hope  to  find  is  traces  of  our  process.  This 
is  why  the  writer  feels  unable  to  attach  much  weight  to  -exact  timing, 
particularly  to  lead  or  lag  of  variations  in  price  level  as  compared  with 
other  elements  of  the  cyclical  process.  They  would  have  to  be  very  con- 
sistent or  very  considerable  to  be  of  real  significance  in  the  circumstances 
under  which  it  is  the  economist's  lot  to  work. 

Expectations  from  the  pure  model  are  so  definite  as  to  make  it  super- 
fluous to  elaborate  them  beyond  what  has  been  said  in  Chap.  IV.  Price 
level  should  rise  in  prosperity — under  the  pressure  of  credit  creation, 
which,  under  the  conditions  embodied  in  the  pure  model,  would  not  be 
compensated  either  by  an  increase  in  output  or  by  any  fall  in  "velocity*' 
— and  fall  in  the  downgrade — under  the  pressure  of  autodeflation  and  of 
increase  in  output — more  than  it  had  risen  in  the  preceding  prosperity.  We 
also  know  that  introduction  of  additional  facts  by  means  of  successive 
approximations  does,  indeed,  tone  down,  but  does  not  reverse,  these 
expectations.  Existence  of  unemployed  resources  in  the  neighborhood 
of  equilibrium  is  one  of  these  facts.  But  the  most  important  difference 
made  by  the  second  approximation — the  substitution  for  the  two-phase 
of  a  four-phase  cycle — adds  the  expectation  that  the  price  level  will  go  on 
falling  in  depression  and  that  this  fall  should  be  corrected  in  recovery. 
This  does  not  mean,  however,  that  recovery  will  carry  the  price  level 
exactly  to  the  figure  at  which  it  stood  in  the  neighborhood  of  incipient 

1  The  writer  has  not  been  able  to  go  into  international  relations — which,  from  the  stand- 
point of  each  individual  country,  constitute  external  factors — to  the  extent  required.  In 
the  matter  of  price  levels  this  lacuna  is  particularly  serious.  For  many  indices  (the  Sauer- 
beck and  the  Soetbeer  indices  are  outstanding  examples;  only  Dr.  Necco's  Italian  index  is 
still  worse)  covariation  is  tautological,  because  they  are  dominated  by  the  great  articles  of 
world  trade.  But  those  articles  exert  their  influence  on  any  index  and  while,  from  one 
standpoint,  this  is  as  it  should  be,  seeing  that  price  levels  are  among  the  most  important 
conductors  of  cyclical  effects,  from  another  standpoint  it  reduces  their  value  as  cyclical 
symptoms  still  further.  It  also  invests  them  with  a  causal  significance  that  they  do  not 
otherwise  have.  The  extent  to  which  levels  have  moved  in  step  in  our  three  countries  is 
best  realized  by  inspecting  the  chart  displaying  rates  of  change,  although  some  readers 
will  find  the  differences  in  behavior  still  more  interesting.  International  comparison 
meets,  however,  with  all  but  insuperable  difficulties.  Professor  Bowley's  work  in  this 
field  (London  and  Cambridge  Economic  Service)  should  be  mentioned.  Finally,  the  writer 
has  never  been  able  to  do  much  with  the  concept  of  a  world's  price  level.  Mention  is  due, 
however,  to  the  work  of  Professor  W.  Gehlhoff,  Die  allgemeine  Preisbewegung,  1890-1918, 
vol.  149,  Part  1  of  the  Schriften  des  Vereins  fUr  Sozialpolitik. 


THE  PRICE  LEVEL  468 

depression.  Even  depression  may,  but  recovery  always  does,  continue 
the  work  of  recession  by  increasing  output.  In  longer  cycles  also  growth 
asserts  itself.  Hence  the  equilibrium  level  that  will  be  reached  by  the 
detour  of  depression  and  revival,  will  in  general  be  lower  than  the 
level  that  obtained  immediately  before  the  system  embarked  upon  it. 
It  should  also  be  observed  that,  depression  being  essentially  erratic, 
it  is  in  each  case  a  question  of  fact  how  much  there  is  for  recovery  to 
correct. 

In  the  long  swing  of  the  Kondratieff,  in  particular,  short-run  fluctua- 
tions such  as  are  caused  by  panics  and  spirals  play  so  small  a  role  that 
there  is  not  the  same  reason  to  expect  a  rise  in  price  level — at  least  in  its 
absolute  values  as  distinguished  from  rates  of  change — during  a  Kondra- 
tieff  recovery  as  there  is  to  expect  it  for  the  recovery  phases  of  the  shorter 
cycles.  The  third  approximation,  which  introduced  the  three-cycle 
schema,  affects  expectations  because  of  the  phenomena  of  interference 
that  henceforth  complicate  the  picture.  Any  given  phase  of  any  given 
cycle  then  comes  under  the  influence  of  the  simultaneous  phases  of  the 
others  and  may  be  entirely  blotted  out  or  even  reversed  by  them.  It  is 
important  to  bear  this  in  mind.  For  while  some  economists,  particularly 
those  who  hold  monetary  theories  of  the  cycle,  may  think  our  expecta- 
tions too  obvious  to  be  worth  stating,  others  deny  that  they  are  borne  out 
by  facts.  And  the  instances  to  which  these  may  point  are  mostly — 
though  not  all  of  them — attributable  to  the  neglect  of  Kondratieff 
effects.1 

Charts  IV,  V,  VI,  and  VII  are  presented  in  order  to  show  how  far  facts 
conform  to  those  expectations.  Chart  IV  displays  the  rates  of  change  of 
price  level  in  our  three  countries,  similarities  between  which  are  as  inter- 
esting as  are  the  differences.  On  the  other  three  charts  the  reader  finds 
the  graphs  (on  a  logarithmic  scale)  of  the  price  levels — the  indices  of 
wholesale  prices — themselves,  together  with  the  graphs  of  indices  of  out- 
put, of  what  we  take  to  represent  the  circulating  medium,  and  of  certain 
rates  of  interest,  which  will  be  discussed  in  the  chapters  that  are  to 
follow.  In  the  workshop  of  the  writer  a  habit  has  grown  up  of  referring  to 
these  charts  as  Pulse  Charts.  The  reason  for  this  is  obvious.  Little 
though  the  writer  thinks  of  the  explanatory  value  of  aggregative  theory, 
and  far  though  he  is  from  claiming  barometric  value  for  the  four  con- 
stituents of  these  charts,  they  nevertheless  give  a  rough  picture  of  the 
economic  process  in  time  and,  in  a  sense,  sum  up  what  we  have  to  account 
for  by  our  analysis.  Also  the  picture  is,  as  far  as  it  goes,  complete:  none 
of  the  four  constituent  curves  could  be  left  out,  but  no  other  is  logically 

1  The  above  refers  only  to  denial  of  the  association  of  rising  prices  with  upgrades  and  of 
falling  prices  with  downgrades.  If  those  authors  only  mean  to  discount  the  causal  role 
of  the  price  level,  we  are,  of  course,  largely  in  sympathy  with  them. 


464 


BUSINESS  CYCLES 


g   <^ 

g  3 


S  -g 
I 


1 


ooopoooo     o  l' 

•—  •       CJ       K>       CM       •"  T* 

4*  I          4          I          +         •*•  I 


THE  PRICE  LEVEL 


465 


necessary  in  order  to  convey  what  meaning  there  is  in  the  variations  of 
aggregates  at  all.  If  the  writer  had  to  construct  an  index  of  business 
conditions,  this  is  what  he  would  offer. 

Naturally,  we  shall  first  look  for  the  result  trend  which  our  analysis 
leads  us  to  expect.  For  since  in  the  downgrades  of  all  cycles  the  price 
level  must,  barring  interference  by  opposite  phases  of  other  cycles,,  fall 
more  than  it  rises  in  their  upgrades,  capitalist  evolution  produces  a  long- 
run  (or  "secular")  tendency  of  prices  to  fall.  This  downward  result 


17851790     1800      1810       1820      1830      1840      1850      1860      1870 

CHART  V. — British  prewar  "pulse"  (see  Appendix, 


1890     1900     191015 
p.  1052). 


trend  embodies  the  method  by  which  the  capitalist  mechanism  diffuses 
the  fruits  of  industrial  improvement  over  the  masses  of  the  people,  and 
characterizes  the  specifically  capitalist  "road  to  plenty."  It  is  an 
illogical  method,  no  doubt,  which  records  increase  in  real  income  in  a 
way  that  may  be  likened  to  measuring  the  growth  of  a  child  by  leaving 
the  number  of  inches  constant  and  increasing  the  size  of  the  inches 
instead.  And  other  methods  of  social  accounting  could  be  devised  which 
might  achieve  the  same  results  without  each  time  creating  the  danger  of 
the  system's  sliding  off  into  depression.  Experience  tends  to  show, 
however,  that  neither  capitalism  itself  nor  the  social  institutions  asso- 


BUSINESS  CYCLES 


elated  with  it,  democracy  among  them,  can  work  efficiently  and  with 
comparative  smoothness  except  on  a  falling  trend  in  prices.1 

The  price-level  curves  in  charts  V,  VI,  and  VII  display,  in  fact,  a 
falling  descriptive  trend.2     But  although  this  descriptive  trend  is  what 


1830    35     1840  45    1850    55     I860   65     1870   75     I860  85     1890    95    1900   05    1910  15 
CHART  VI. — United  States  prewar  "pulse'*  (see  Appendix,  p.  1053). 

is  left  in  our  figures  of  the  result  trend,  the  former  does  not  of  course 
render  the  latter.     The  two  differ  by  the  effects  of  external  factors. 

1  That  statement  rests  on  the  opinion  that  all  alternatives  that  are  politically  feasible 
carry  with  them  other  effects  which  in  one  way  or  another  tend  to  upset  the  working  of  the 
system.     It  should  again  be  remembered,  however,  that  what  has  been  said  does  not  apply 
to  the  fall  of  prices  in  depression. 

2  In  retail  prices,  particularly  if  we  include  rent,  it  is  much  less  pronounced.     But  no 
index,  however  constructed,  has,  as  far  as  the  writer  knows,  succeeded  in  effacing  or  revers- 
ing it.     It  is  of  course  quite  incorrect  to  include  wages. 


THE  PRICE  LEVEL 


467 


Government  deflations  have  never  gone  far  enough  to  counter-balance 
the  corresponding  government  inflations,  which  in  our  countries  and 
period  have  been  mainly,  though  not  exclusively,  those  of  the  Napoleonic 
time  and  of  the  American  Civil  War.  Demonetization  of  silver,  on  the 
one  hand,  and  the  impact  of  Australian,  Californian,  and  South  African 
gold  discoveries,  on  the  other,  must  have  exerted  some  influence.  .The 
spread  of  deposit  banking  acted  in  the  same  direction  as  the  gold  dis- 
coveries. Protectionism  and  immigration  of  capital  also  have,  in  some 
cases,  been  responsible  for  rising  prices,  or  have  prevented  a  fall  that 
would  have  occurred  otherwise.  Theoretically  the  descriptive  trend  of 
the  sum  total  of  money  incomes  should  measure  the  net  influence  of  all 


65  1870          75  1880          85  1890          95  1900          05  1910     13 

CHART  VII. — German  prewar  "pulse"  (see  Appendix,  p.  1054). 

these  external  factors.  Unfortunately,  however,  the  evolutionary 
process  itself  tends  to  bring  about  not  only  the  recurrent  expansions  and 
contractions  of  deposits  that  are  described  by  our  pure  model,  but  also 
those  lasting  expansions  which  we  took  into  account  in  our  successive 
approximations  and  in  consequence  of  which  money  incomes  will  display 
a  long-time  tendency  to  increase.  Hence,  these  expansions  become  part 
of  our  process,  and  the  method  alluded  to,  therefore,  cannot  be  expected 
to  serve  as  a  means  of  separating  the  result  trend  from  the  effects  of  other 
factors.  Moreover,  there  are  other  reasons  for  this.  Structural  changes  in 
habits  of  payment  are  also  induced  by  our  process  itself.  Increase  in  the 
supply  of  monetary  gold  cannot  be  considered  as  a  wholly  independent 
factor.  Government  inflations  and 'deflations  and  the  policies  of  which 
they  are  elements  have  many  effects  on  the  economic  organism  besides 


468 


BUSINESS  CYCLES 


influencing  incomes.  A  very  interesting  research  program  suggests 
itself  here.  It  is,  however,  safe  to  say  that,  for  our  countries  at  least, 
the  net  effect  of  external  factors  has  been  to  counteract  and  not  to  inten- 
sify the  influence  of  the  result  trend. 

It  stands  to  reason  that  external  factors  affect  not  only  the  result 
trend  but  also  the  cyclical  behavior  of  price-level  series.  Moreover,  the 
facts  that  our  figures  do  not  express  at  all  accurately  the  level  of  prices 
in  our  sense  and  that  they  even  do  not  measure  correctly  what  they  could 
be  expected  to  measure  are  likely  to  make  themselves  felt  still  more  in 
the  case  of  cycles  than  they  do  in  the  case  of  the  secular  tendency. 
Finally,  there  is  always  so  much  "slack"  and  underemployment  in  the 


1780     1790     1800      1810      1820      1830      1840     1850     1860      1870      1880      1890     1900     1910 
CHART  VTII. — Nine  year  moving  averages  of  price  indices  (see  Appendix  p.  1054). 

system  and  so  much  room  for  the  play  of  a  multitude  of  factors  that  the 
effects  on  the  price  level  of  each  cyclical  phase  may  be  slow  to  emerge  and 
can  easily  be  drowned  in  the  effects  of  the  succeeding  one.  It  should  be 
equally  clear  that  this  does  not  invalidate  our  analysis. 

In  the  case  of  the  Kondratieff,  some  readers  might  think  it  a  waste  of 
space  to  prove  the  existence  of  those  protracted  periods  of  rising  and  of 
falling  prices  which  stand  out  clearly  enough  and  are  seen  at  first  glance 
to  correspond  roughly  with  the  upgrades  and  downgrades  of  that  long 
wave.  In  fact,  all  we  need  to  do  in  order  to  prove  this  is  to  look  at  the 
pulse  charts  or  to  take  a  nine-year  moving  average  of  American,  British, 
and  German  indices  of  prices  at  wholesale  (see  Chart  VIII).  It  cannot 
be  too  often  repeated,  however,  that  mere  shifts  in  the  price  system, 


THE  PRICE  LEVEL 


469 


though,  they  do  not  per  se  influence  our  price  level,  do  influence  the  indices 
we  have.  Hence  individual  prices  may,  besides  their  legitimate  influence, 
acquire  an  illegitimate  one  as  well. 

We  see  fairly  smooth  wavelike  lines  that  display  two  units  of  about 
equal  length  and  the  beginning  of  a  third.  Chart  IX,  which  gives  the 
variations  of  American  wholesale  prices  treated  by  a  simplified  freehand 
adaptation  of  Professor  Frisch's  method,  discussed  in  Chap.  V,  presents  a 
similar  picture.  Recalling  the  testimony  of  industrial  history,  we  shall 
associate  the  first  unit  with  the  processes  usually  referred  to  as  the  indus- 
trial revolution.  But  the  height  and  location  of  the  maxima,  being 


A  3  YEAR  ON  A  2  YEAR  MOVING 
AVERAGE  (GRAPHICALLY 
~     DETERMINED) 


CURVE  THROUGH 
INFLECTION 
POINTS  OF  50        \\ 

YEAR  CYCLE      -\\ 


:ffiVE  THROUGH  INFLECTION 
POINTS  OF  22  YEAR  CYCLE 


1800    1810     1820      1830      1840     1850      1860      1870     1880     1890     1900      1919      1920 
CHART  IX.— (See  Appendix  p.  1054.) 


obviously  conditioned  by  the  Napoleonic  Wars,  prove  nothing.  The 
English  maximum  (Silberling  index)  occurs  in  1814,1  after  which  "defla- 
tion" due  to  normalization  of  government  finance,  superimposing  itself  on 

1  Mrs.  Gilboy  (Review  of  Economic  Statistics,  1936)  gives  1818,  but  since  she  uses  harvest 
years,  there  is  really  no  difference.  Like  Jevons  before  us,  we  should  have  expected  the 
maximum  to  occur  earlier,  the  more  so  because  English  war  finance,  with  relapses,  grew 
steadily  sounder  after  the  turn  of  the  century,  though  Mr.  Silberling's  figures  do  not  quite 
agree  with  this.  Dr.  E.  B.  Schumpeter  (Review  of  Economic  Statistics,  Feb.  1938)  in  fact 
established  that  prices  of  domestic  commodities  fell  from  1800,  which  would  agree  well 
with  our  cyclical  schema.  It  must  be  remembered,  however,  that  the  influx  of  precious 
metals  from  Spanish  South  America,  which  before  the  wars  of  the  French  Revolution  was 
over  7  million  pounds  per  year,  began  to  fall  later  on.  It  was  about  5  millions  in  1825  and 
only  4  millions  in  1829. 


470  BUSINESS  CYCLES 

the  autodeflation  we  should  theoretically  expect,  possibly  blurs  the  picture 
for  about  half  a  dozen  years.  We  cannot  eliminate  this  disturbance  by 
using  gold  instead  of  paper  prices,  for  gold  prices  also  are  influenced  by 
inflation  and  deflation  and  by  the  events  of  which  inflation  and  deflation 
are  but  the  monetary  garb.  The  failure  of  our  figures  for  price  level  to 
rise  in  the  eighties  of  the  eighteenth  century  may,  in  the  case  of  England, 
be  due  to  imperfections  of  our  material.  New  evidence  of  increase  in 
prices  of  many  important  commodities  points  in  that  direction.  In 
America,  the  beginning  of  the  Kondratieff  is,  as  we  have  seen,  doubtful, 
owing  to  the  Revolution  and  its  aftermath.1 

The  third  and  smaller  hump  that  we  observe  in  the  chart  of  nine- 
years  moving  averages  presents  more  delicate  problems.  The  occurrence 
of  such  a  fluctuation  is  not  astonishing  in  itself  and  proves  nothing  against 
the  three-cycle  schema,  because  it  may  be  but  an  effect  of  superposition 
(see  Chart  I  and  the  top  curve  in  Professor  Fisher's  chart  in  Business 
Cycles  as  Facts  or  Tendencies,  p.  6,  in  the  volume  in  honor  of  Professor  C. 
A.  Verrijn  Stuart).  But  the  amplitude  is  unexpectedly  great  for  a  Juglar 
belonging  to  the  recovery  phase  of  a  Kondratieff.  Moreover,  price  levels 
continued  to  fall  far  into  the  forties,  according  to  some  indices  much  beyond 
our  date  for  the  rise  of  the  prosperity  of  the  second  Kondratieff.  We 
have  met  the  same  difficulty  in  our  historical  discussion.2  Of  course, 
it  may  mean  no  more  than  that  our  schema  is  faultily  designed.  This 
may  well  be  so.  But,  as  stated  in  Chap.  VI,  the  writer  feels  inclined 
to  believe  that  "reckless"  banking  practice,  which  undoubtedly  was  an 
outstanding  feature  of  the  thirties  in  both  America  and  England,  inten- 
sified the  prosperity  phase  of  the  last  Juglar  of  that  Kondratieff  and 
induced  a  wave  of  speculation  of  unusual  violence  which,  in  turn,  accounts 
for  the  depression  of  unusual  severity,  and  for  underemployment  at  the 
beginning  of  the  second  Kondratieff.  This  explanation  seems  partic- 
ularly plausible  in  its  application  to  the  price  level,  while  political  troubles 

1  Mr.  H.  M.  Stoker  speaks  of  a  postwar  depression  that  lasted  for  eight  years  (Memoran- 
dum 142,  Cornell  Agricultural  Experiment  Station,  1932,  p.  £04).     This  hardly  accords  with 
historical  evidence  (see  Professor  W.  B.  Smith's  comments  in  Smith  and  Cole,  Fluctuations 
in  American  Business,  1790-1860,  1935,  p.  12).     But  it  is  true  that  existing  figures  do  not 
indicate  a  rise  in  price  level  before  1792.     Miss  Crandall's  "special"  index  of  Boston  prices 
does  rise  from  1788,  but  it  consists  only  of  molasses,  rum,  and  fish.     The  situation  was, 
as  we  have  seen,  prosperous  in  1790,  and  became  violently  booming  immediately  afterward 
(see  Chap.  VI,  Sec.  D).     The  Hamilton  policy,  though  favorable  to  prosperity,  was  not 
favorable  to  an  increase  in  prices.     Also,  the  virgin  environment  offered  plenty  of  under- 
employed resources. 

2  Therefore,  Kondratieff  himself  dated  the  second  long  wave  from  1849.     Our  view  is, 
however,  supported  by  the  authority  of  Spiethoff,  who  has  an  Aufschwungsspanne  1842-1873 
on  the  criterion,  perfectly  valid  from  our  standpoint,  that  years  of  prosperity  (he  counts  21) 
predominate  over  years  of  depression  (10). 


THE  PRICE  LEVEL  471 

would,  as  has  been  pointed  out  by  Sauerbeck  himself  (Journal  of  the  Royal 
Statistical  Society,  1886,  p.  648 l),  no  less  plausibly  account  for  the  further 
fall,  1848  to  1851,  which  is,  however,  perfectly  regular  within  the  course 
of  the  Juglar.  Moreover,  in  England,  where  the  fall  of  the  price  level  is 
most  obvious,  introduction  of  free  trade  must  have  had  some  effect  on 
prices,  and  Peel's  Act  may  have  exerted  some  restrictive  effects  on  credit 
creation,  as  in  fact  it  was  intended  to  have.  If  there  is  anything  in  this, 
we  must  conclude  that  we  have  simply  a  case  of  a  primary  cyclical  element 
failing  to  behave  according  to  expectation  because  of  a  combination  of 
counteracting  circumstances.  That  this  did  not  prevent  a  prosperity 
phase  in  our  sense  from  starting  and  running  its  course  is  clear  in  any 
case :  the  English  railway  mania  testifies  to  that  with  no  uncertain  voice. 

The  behavior  of  certain  indices  seems  to  lend  some  support  to  this 
interpretation.  In  Germany,  where  the  hump  of  the  thirties,  though  not 
absent,  is  much  less  in  evidence  than  in  England  and  America,  prices 
began  to  rise  in  1845  (see  the  Berlin  Institut's  new  index).  In  France 
there  was  a  sharp  upturn  of  the  price  level,  at  least  from  the  beginning  of 
1844,  possibly  from  the  end  of  1843.  For  the  United  States,  Professor 
Cole's  38-commodity  index  (Review  of  Economic  Statistics  for  April  1926; 
1834-1842  =  100)  gives  a  maximum  for  1836  (130)  and  a  minimum  for 
1842  (72),  after  which  a  rise  set  in  that  lasted  up  to  the  point  where  we 
locate  the  turn  of  that  Juglar  (1847).  Professors  Warren  and  Pearson's 
all-commodity  index  reaches  its  trough  in  1843  and  then  begins  to  rise.  So 
do  their  indices  of  metal  and  metal  products  and  of  building  material. 
It  does  not  seem  unreasonable  to  describe  this  as  a  tendency  toward 
increase,  dwarfed  by  the  circumstances  referred  to  above. 

Recalling  what  has  been  said  on  the  subject  in  Chap.  VII,  we  find 
that  in  all  three  countries  (also  in  France,  according  to  the  index  of  the 
Statistique  Generale  de  France)  a  peak  occurs  in  the  middle  of  the  fifties 
that  may  be  identified  as  a  perfectly  regular  Kondratieff  turning  point 


Dates  of  Minimum  for  Sauerbeck  Group  Indexes 

All  commodities .      .  .  1849,  1851  (equal) 

Total  materials 1848,  1849  (equal) 

Food 1851 

Corn,  etc..  .  1851 

Animal  foods 1850 

Sugar,  coffee,  tea 1848 

Minerals 1851 

Textiles 1848 

Sundry  materials 1849 

See  Journal  of  the  Royal  Statistical  Society,  1886,  p.  648,     Ten-year  moving  average  touches 
its  minimum  in  1848.     However,  the  Sauerbeck  index  is  no  ideal  guide. 


47£  BUSINESS  CYCLES 

But  the  fall,  which  we  should  then  expect  to  continue,  with  interruptions 
due  to  the  prosperities  of  the  shorter  cycles,  for  at  least  the  whole  of  the 
Kondratieff  recession  and  depression  phases,  was  everywhere,  except  in 
France,  checked  about  1858.  To  about  1863  the  Juglar  recovery  and  pros- 
perity account  for  that,  when  the  German  price  level  in  fact  begins  to  fall. 
That  turning  point  occurs  in  England  one  year  later,  and  in  the  United 
States  (if  we  take  gold  prices)  two  years  later.  This  and  the  violence — 
not  the  fact — of  the  rise  in  prices,  which  in  England"  and  Germany 
occurred  from  1870  to  1873 — both  gold  and  currency  prices  moved  up  but 
moderately  in  this  country  in  1871  and  the  first  half  of  187& — afford  the 
only  occasion  for  invoking  the  influence  of  gold  production  in  explanation 
of  movements  contrary  to  expectation,  although  on  others  that  influence 
may  have  intensified  tendencies  independently  induced  by  different 
factors.  We  shall  understand,  however,  that  economists  who  put  their 
trust  in  the  formal  properties  of  a  graph  will  look  upon  the  whole 
segment  from  about  1849  to  1873  as  a  long-time  upgrade.  But  since  we 
can  easily  account  for  that  peak — which,  to  repeat,  is  irregular  in  nothing 
but  height — and  since  the  reaction  was  so  prompt  and  so  severe,  the 
thesis  that  the  underlying  tendency  was  downward  from  the  middle 
fifties  is  perhaps  not  indefensible.  If  so,  the  consequence  that  follows 
for  the  value  of  formal  methods  of  analysis  is  truly  discouraging.  Price 
levels  then  continued  to  fall,  not  only  through  the  depression  but  also 
through  the  recovery  of  the  Kondratieff,  though  for  all  our  countries 
it  is  possible  to  speak  of  a  fall  at  a  decreasing  rate,  which  in  the  case  of 
Germany  almost  vanishes  by  1886.  Thus  Kondratieff  recovery  again 
failed  to  bring  about  a  recovery  in  prices.  The  reasons  have  been  men- 
tioned at  the  beginning  of  this  section.  The  rise  of  the  third  Kondratieff 
is  clearly  marked  and  particularly  normal  in  this  country  (1897). 

We  may  sum  up  by  saying  that  the  great  waves  of  economic  change 
recorded  by  history  show  in  the  behavior  of  the  price  level,  but  that  the 
association  is  so  imperfect  as  to  make  it  highly  unreliable  for  purposes 
of  diagnosis  or  prognosis.  Since  existence  and  adequacy  of  the  disturb- 
ances that  we  hold  responsible  for  that  imperfection  can  in  each  case 
be  established  from  independent  historical  evidence,  the  fact  should  not 
be  recorded  against  our  model.  Among  them,  monetary  disorders,  which 
in  particular  account  for  the  outstanding  peaks,  are  by  far  the  most  impor- 
tant. But  again  the  question  whether  we  have  done  justice  to  the  gold 
factor  crosses  our  way.  The  problem  of  its  possible  role  in  the  causation 
of  business  cycles  having  been  dealt  with  in  Chap.  IV,  Sec.  E  and  Chap.  VI, 
Sec.  C,  it  remains  to  touch  upon  the  problem  of  its  causal  importance  for 
the  price  level. 

Professor  Gustaf  Cassel,  whose  advocacy  of  the  time-honored  view 
that  the  price  level  is  dominated  by  gold  production  in  relation  to  output 


THE  PRICE  LEVEL  473 

or,  more  generally,  economic  activity1  has  been  instrumental  in  starting 
the  discussion  anew,  compared  world  stock  of  gold  for  1850  and  1910,  in 
which  years  the  Sauerbeck  index  stood  at  approximately  the  same  figure, 
concluded  that  the  average  increase  in  gold  stocks  between  those  years 
(2.8  per  cent)  was  "normal"  in  the  sense  that  it  was  sufficient  to  keep  the 
price  level  constant,  and  compared  variations  of  the  ratio  of  the  actual  to 
the  normal  quantity  of  gold  with  the  variations  in  price  level  that  had 
occurred,  extrapolating  to  1800.  Mr.  Joseph  Kitchins  corrected  figures, 
took  account  of  Indian,  Chinese,  and  Egyptian  hoarding,  restricted  gold 
stocks  to  monetary  gold  stocks,  and  arrived  at  3.1  per  cent  as  the  normal 
yearly  increment.  Fit  was  greatly  improved  and  the  obviously  desirable 
precedence  of  variations  in  that  ratio,  as  against  variations  in  price  level, 
was  not  wanting.  Even  so,  the  method  not  unnaturally  met  with  adverse 
criticism.2  Besides,  it  might  be  urged  that  the  growth  of  deposit  banking 
must  have  alleviated  the  "scarcity"  of  gold  after  1873  and  extension  of  the 
area  of  the  gold  standard,  the  plethora  of  gold  after  1896.  Much  further 
research  will  have  to  be  done  before  we  can  speak  with  confidence  on  these 
matters;  but  for  our  purpose  we  need  not  go  into  them.  We  will  grant 
for  argument's  sake  more  than  we  believe  should  be  granted  and  accept 
as  it  stands  the  evidence  offered  for  the  effects  of  gold  on  price  level.  It 
does  not  follow  that  the  Kondratieff  wave  in  price  level  is  simply  due  to  the 
variations  in  gold  production.  On  the  contrary,  it  is  clear — since  accord- 
ing to  that  theory  price  level  is  the  result  of  variation  of  monetary  gold 
stocks  (which,  let  us  note  in  passing,  are  still  more  a  function  of  business 
situations  than  total  gold  stocks)  and  output  of  commodities,  and  since 
variation  in  the  latter  result,  in  turn,  from  the  working  of  our  process — 
that  whatever  the  behavior  of  gold,  unless  it  should  happen  to  be  exactly 
compensatory,  the  finger  prints  of  the  Kondratieff  must  show  on  price- 
level  graphs,  although  more  or  less  blurred  by  gold  production. 

1  Having  already  presented  his  method  before  the  war,  he  substantially  repeated  his 
argument  in  the  League  of  Nations  Rapport  Provisoire  de  la  Delegation  de  I' Or  du  C omit 6 
Financier,  annexe  X.    Mr.  J.  Kitchin's  work  on  the  subject  was  also  done  for  the  same 
delegation. 

2  The  reader  finds  an  exposition  and  defence  of  the  method  in  W.  Woytinsky,  Inter- 
nationale Hebung  der  Preise  als  Ausweg  aus  der  Krise,  1931,  and  a  good  presentation  of  the 
case  against  it  in  Mr.  J.  T.  Phinney's  article  Gold  Production  and  the  Price  Level.     The 
Cassel  3  %  Estimate,  Quarterly  Journal  of  Economics,  1933.     See  also  Mr.  Tucker's  paper  on 
Gold  and  Prices,  Review  of  Economic  Statistics,  1934.     Fit  is  even  better  than  the  sponsors 
of  the  method  claim,  for  they  fail  to  take  account  of  the  important  role  silver  played  in  the 
first  half  of  the  century  where,  over  the  interpolated  range,  the  fit  is  less  satisfactory. 
Overlooking  this,  Mr.  Woytinsky  tries  to  improve  the  fit  by  assuming  a  smaller  rate  of 
progress  during  that  time,  in  order  to  pull  down  the  curve  of  the  ratio  between  actual  and 
normal  gold  supply  until  it  goes  through  the  curve  of  the  price  level.    But  there  is  as  little 
necessity  for  this  "correction**  as  there  is  warrant  for  it. 


474 


BUSINESS  CYCLES 


In  spite  of  all  the  objections  that  may  justly  be  raised  against  the  use 
of  moving  averages,  we  will  take  the  deviations  from  our  nine-year  moving 
averages  in  order  to  get  some  idea  of  the  Juglars.  Chart  X  exhibits  the 
result.  Although  with  characteristic  differences  in  amplitude,  they  seem 
to  the  writer  to  stand  out  rather  well  for  all  three  countries.  There  are, 
in  England  and  the  United  States  (but  we  may  with  practical  certainty 
assert  the  same  thing  for  Germany),  six  of  them  to  the  only  complete 
Kondratieff  covered  by  the  chart.  The  reader  should  -mark  them  off 
and  form  an  opinion  about  the  question  how  far  there  would  be  any  sense 


Per  Cent 
+20 

+10 

0 

-10 
+30 
+20 
+  10 

0 
-10 

+20 

+  10 

0 

-10 
-20 
-30 


BRITAIN 


^ 


A, 


\r 


V 


/Wr 


w/ 


V 


UNITED  STATES 


V 


V 


1780     1790      1800     1810      1820      1830     1840      I860     I860      1870     1880     1890     1900     1910 
CHABT  X. — Percentage  deviations  of  price  indices  from  their  nine  year  moving  averages 

(see  Appendix  p.  1055). 

in  trying  to  measure  average  amplitudes — bearing  in  mind  the  Kondra- 
tieff phases  from  which  they  rise  and  to  which  they  have,  in  each  case, 
to  fall — and  to  base  conclusions  on  such  averages.  In  passing,  we  will 
advert  to  Mr.  Philip  Wright's  experiment  (Moore's  Economic  Cycles, 
Quarterly  Journal  of  Economics,  1914-1915,  p.  635),  which  consisted  in 
correlating  successive  items.  Correlation  was  found  to  be  highest  for 
items  distant  from  each  other  by  9  or  10  years.  The  coefficients  are  low — 
.40  for  9  years  and  .35  for  10  years — and  the  procedure  is  otherwise  open 
to  criticism.  But  the  result  accords  well  with  our  findings.  Also,  it 
should  be  observed  that  some  of  the  objections  to  applying  correlation 
analysis  to  time  series  do  not  hold  in  this  case.  Professor  E.  B.  Wilson's 
periodogram  of  prices  shows  only  an  unsatisfactory  hump  for  a  109-month 


THE  PRICE  LEVEL,  475 

period,  but  it  may  be  significant  that  other  humps  also  occur  at  periods 
of  about  twice  and  three  times  that  length.1 

Charts  X  and  IV  may  be  consulted  in  order  to  get  an  impression  of  the 
Kitchins.  No  doubt  is  possible  as  to  the  presence  of  shorter  fluctuations 
within  the  Juglars.  It  seems  fairly  clear  also  that  there  are  about  three 
of  them  to  each  Juglar.  They  are  highly  irregular  in  amplitude,  some- 
times showing  by  a  mere  kink  in  the  curve.  This  can  however  always  be 
explained  by  the  influence  of  underlying  cycles,  which — as  mentioned 
above — may  go  so  far  as  to  iron  them  out.  Their  period  is  less  irregular 
and  does  not  stray  far  from  the  average. 

E.  Group  Prices. — An  index  of  a  family  of  prices  differs,  toto  caelo, 
from  what  we  have  been  discussing  in  the  preceding  sections.  There  is 
no  objection  of  course  to  calling  it  a  level,  but  it  is  not  a  parameter  char- 
acteristic of  the  system  as  such  or  of  the  monetary  ligamen.  On  the  one 
hand  it  is  no  "real  thing"  existing  irrespective  of  the  doings  of  statis- 
ticians :  it  owes  its  existence,  not  only  its  presentation,  to  them.  On  the 
other  hand,  it  remains  a  price,  subject  (if  the  group  is  not  too  big2)  to 
the  ordinary  logic  of  prices,  and  should  simply  be  called  a  composite  price 
or  a  composite  price  relative,  as  the  case  may  be.  In  the  most  favorable 
instances  the  matter  involves  no  other  problems  than  are  always  implied 
in  speaking  of  "the"  price  of  a  not  perfectly  homogeneous  commodity  or, 
say,  the  composite  price  of  finished  steel.  In  other  instances  the  object 
is  to  combine  commodities  the  prices  of  which  display  some  feature  com- 
mon to  all  of  them,  without  attempting  to  find  anything  that  will  bear 
interpretation  as  a  composite  price;  we  shall  look  upon  indices  of  group 
prices  in  this  light. 

One  type  of  case  may  be  instanced  by  the  familiar  indices  of  prices  of 
textiles,  metals  and  metal  products  or,  if  the  relationship  is  on  the  side 
of  demand,  by  indices  of  prices  of  building  materials.  Another  type 
consists  of  groups  such  as  the  group  of  Sensitive  Prices,3  so  well  known 
to  students  of  forecasting  in  this  country  and  in  Germany.  This  is 
merely  an  assemblage  of  prices  empirically  found  to  fluctuate  more 

1  The  periods  of  the  Juglars  do  not  seem  to  be  affected  by  the  Kondratieff  phases;  but 
duration  of  the  Juglar  phases  may  be,  i.e.,  prosperities  may  be  not  only  more  marked  but 
also  longer  in  Kondratieff  prosperities,  than  in  Kondratieff  downgrades.     This  is  what  has 
been  observed  by  Professor  Spiethoff.     The  writer  prefers  to  say  that  in  the  former  cases 
recessions  look  more  like  prosperities  and  in  the  latter,  more  like  depressions,  and  not  to 
stress  difference  in  duration.     But  this  is  little  more  than  a  different  way  of  expressing  the 
same  thing. 

2  If  it  is  too  big,  the  Marshallian  supply-and-demand  apparatus  loses  its  meaning. 
We  may  still  speak  of  a  composite  price,  but  only  the  roughest  of  popular  meanings  attaches 
then  to  the  concepts  of  supply  and  demand,  and  they  will  no  longer  bear  modern  refinements. 

3  See,  for  instance,  Professor  W.  M.  Persons'  index  in  Forecasting  Business  Cycles, 
p.  93,  et  seq.,  also  the  index  of  the  Institut  filr  Konjunkturforschung. 


476  BUSINESS  CYCLES 

violently  than  others.  It  serves  a  useful  purpose,  but  since  for  us  the 
wholesale  price  index  is  quite  sensitive  enough,  we  may  dismiss  this  group 
with  the  remark  that  its  divergence  from  the  general  index  gives  what  so 
far  has  been  the  relatively  most  successful  measure  of  cyclical  variation 
of  price  dispersion.  Price  indices  of  durable  and  transient,  of  domestic 
and  foreign,  of  monopolized  and  not  monopolized,  of  raw  and  processed, 
of  basic  and  other  commodities,  of  foods  and  nonf oods,  are  instances  of  the 
application  of  what  may  fairly  be  called  a  distinct  method  of  analysis. 
Regional  group  prices  are  really  levels.  Then  there  is  the  great  division 
between  producers'  and  consumers'  goods  prices,  which  comes  near  to 
defining  entire  spheres  of  the  monetary  process,  provided  we  look  on  both 
the  producers'  and  the  consumers'  goods  prices  included  as  random 
samples. 

Finally,  we  might  combine,  with  suitable  weights,  an  index  of  prices 
of  raw  materials  with,  first,  a  composite  of  an  index  of  prices  of  equipment 
and  of  building  costs,  and  second,  an  index  of  wage  rates  corrected  for 
changes  in  product  per  man-hour.  Thus  we  could  get  an  index  of  costs 
and  compare  it  with  an  index  of  prices  of  finished  products.  But  unless 
considerable  means  were  invested  in  this  project,  results  would  be  exposed 
to  such  serious  errors  that  we  could  not  trust  them  beyond  what  we  may 
in  any  case  infer  from  indices  of  prices  of  raw  materials,  wages,  and  so  on, 
taken  separately.1  More  progress  has  been  made,  since  the  war  at  least, 
in  the  construction  of  indices  of  the  prices  of  wage  goods — the  cost-of- 
living  indices. 

Suitably  chosen  group  prices  might  be  expected  to  show,  in  the  course 
of  the  cycle,  characteristic  variations  relatively  to  each  other.  So  they 

1  The  problem  is  different,  of  course,  according  to  whether  it  is  desired  to  build  a  "  social" 
index  of  cost  of  production  (or  cost  of  doing  business  in  general)  or  whether  the  goal  is  the 
more  modest  one  of  deriving  an  index  of  costs  for  a  given  industry  or,  still  more  modestly, 
a  given  concern.  Professor  Mills'  investigations  on  cost  in  his  Economic  Tendencies  in 
the  United  States,  the  cost  indices  of  the  late  G.  T.  Jones  (Increasing  Returns,  posthum- 
ously published,  1983),  and  also  Professor  Mitchell's  pioneer  work  in  his  great  book  of  1918 
must  be  mentioned.  The  theoretical  bases  for  further  advance  have,  it  seems  to  the  writer, 
been  laid  in  Professor  Leontief's  Studies  on  the  Elasticity  of  Supply  (Weltwirtschaftliches 
Archiv  for  January  1982),  which  paper  also  presents  interesting  applications  to  the  iron  and 
steel  industry  of  the  United  States.  An  example  of  an  index  of  cost  for  one  concern  is  given 
in  K.  Ehrke:  Uebererzeugung  in  der  Zementindustrie,  1858-1918  (1983;  the  method  is 
Dr.  Schneider's).  Cost  investigations,  of  course,  abound  since  the  war;  but  they  rarely 
cover  more  than  a  few  years.  Construction  costs  are  a  particularly  bright  spot.  See, 
for  example,  the  index  of  the  Federal  Reserve  Bank  of  New  York  and  that  of  the  American 
Appraisal  Company.  Compare  L.  J.  Chawner,  Construction  Cost  Indexes  as  Influenced  by 
Technological  Change  and  other  Factors,  Journal  of  the  American  Statistical  Association 
for  1935,  particularly,  Chart  V  on  p.  57,  and  W.  D.  Conklin,  Building  Costs  in  the  Business 
Cycle.  These  studies  are  mentioned  here,  although  they  deal  almost  exclusively  with  war 
and  postwar  material,  because  some  of  the  questions  of  principle  involved  can  hardly  be 
dealt  with  on  prewar  material. 


THE  PRICE  LEVEL  477 

do,  of  course.  These  variations  can  be  brought  out  by  dividing  all  of  them 
by  the  index  of  the  general  price  level1  and  can  be  used  for  the  purpose  of 
measuring  certain  types  of  cyclical  price  dispersions  and  disequilibria. 
They  are  of  considerable  importance  for  the  higher  approximations  of  a 
more  refined  analysis  and  for  any  quantitatively  exact  picture  of  cyclical 
situations.  But  if  we  look  at  Chart  XI,  which  is  sufficient  to  illustrate 
the  one  point  we  wish  to  make  and  which  the  reader  can  easily  supplement 
by  other  such  graphs  familiar  to  everyone,  it  is  not  that  aspect  which 
strikes  us  first.  What  stands  out  is,  on  the  contrary,  the  covariation: 
compared  with  it,  all  there  is  of  difference  in  form,  amplitude,  and  period, 
and  of  lag  is  distinctly  secondary.2  Let  us  pause  for  a  moment  to  con- 
sider that  broad  truth  which,  though  only  broadly,  yet  asserts  itself 
consistently. 

From  our  analysis  of  the  cyclical  process  of  evolution  it  follows,  indeed, 
that  cycles  are  not  satisfactorily  described  as  aggregative  movements 
that  leave  structural  relations  within  the  system  untouched.  It  is  of 
the  very  essence  of  that  process  that  it  remodels  the  structure  of  the 
system.  But  it  does  not  follow  that  comparison  of  the  cyclical  behavior 
of  group  prices  must  show  this.  We  will  divide  all  the  groupings  of 
prices  that  have  been  mentioned  above  into  three  classes.  To  the  first 
we  assign  those  group  prices  the  constituents  of  which  are  related  by 
virtue  of  affinity  between  the  commodities — like  group  prices  of  textiles, 
electrical  apparatus,  and  so  on.  To  the  second  belong  group  prices 
formed  from  prices  of  commodities  that  have  in  common  some  character- 
istic of  organization  or  marketing;  prices  in  large-scale  industries  or 
monopolized  industries  afford  examples.  Into  the  third  we  put  group 
prices  built  from  constituents  which  dwell  in  distinct  "stages"  of  the 
economic  process — producers'  and  consumers'  goods  being  a  typical 
example. 

Now,  we  shall  expect  that,  both  within  the  shorter  cycles  and  in  the 
long  run,  group  prices  of  the  first  class  will  behave  differently  according 

1  This  has  been  done  very  often,  particularly  in  order  to  express  variations  in  the  "pur- 
chasing power"  of  the  products  of  agriculture  (the  German  word  Agrarschere  has  no  good 
English  equivalent),  see,  for  example,  Kondratieff,  Die  Preisdynamik  der  industriellen 
und  der  landwirtschaftlichen  Waren,  Archiv  fur  Sozialwissenschaft,  vol.  60.  Mr.  Carl 
Snyder  made  extensive  use  of  that  device  in  his  papers  on  the  Structure  and  Inertia  of 
Prices  (American  Economic  Review  for  June  1934)  and  on  Commodity  Prices  versus  the 
General  Price  Level  (ibid.,  September  1984).  In  the  facts  and  in  some  results  the  writer 
entirely  agrees  with  him.  The  above  may  even  be  considered  as  an  interpretation  of  some 
of  his  findings.  However,  the  obvious  dangers  of  the  method,  particularly  when  applied  to 
groups,  should  not  be  overlooked. 

8  This  would  be  different  if,  following  some  students,  we  took  the  price  of  shares  (as 
represented,  say,  by  the  A-CUTVG  of  the  Harvard  Barometer)  to  mean  price  of  real  capital. 
But  to  the  writer  this  seems  entirely  inadmissible. 


478 


BUSINESS  CYCLES 


to  whether  or  not  they  are  dominated  by  prices  of  industries  that  happen, 
in  each  interval  of  time,  to  be  innovating.  There  are  indeed  many 
qualifications  to  this.  New  commodities  lack,  at  their  start,  a  standard 
of  comparison  by  which  to  characterize  their  relative  behavior.  As  far 


1840   45     1850     55     1860    65      1870    75     1880    85     1890    95      1900    05      1910  13 
CHART  XI. — United  States  group  prices  (see  Appendix  p.  1055). 

as  innovations  occur  in  highly  finished  goods,  the  price  variations  easily 
escape  record,  both  for  statistical  reasons  and  because  such  innovations 
often  result  in  the  offering  of  better  qualities  for  the  same  price.  Even 
where  innovation  does  influence  quoted  price,  the  effect  may  to  a  large 


THE  PRICE  LEVEL  479 

extent  be  lost  by  grouping.  Moreover,  innovation  in  most  cases  also 
acts  on  other  prices  as  well  as  on  the  prices  of  the  innovating  industry, 
and  entrepreneurial  activity  shifts  from  one  industry  to  another,  so 
that  we  must  look  for  its1  influence — particularly  its  long-run  influence 
— to  the  price  level  rather  than  to  group  prices,  which  for  those  reasons 
will  tend  to  "catch  up"  with  each  other.  But  we  may  still  expect 
innovation  to  show  and  so  it  does.  If  the  reader  inspect  Mr.  Snyder's 
charts  (in  the  first  of  the  two  papers  quoted  in  our  footnote,  pp.  189  and 
190)  he  will  not  find  it  difficult  to  associate  the  behavior  of  the  group 
prices  of  chemicals  and  metals  or  (in  the  late  seventies  and  eighties)  of 
fuel  and  lighting  with  innovation,  or  the  behavior  of  the  group  prices  of 
building  materials  and  of  hides  and  leather  with  (comparatively  speaking) 
the  absence  of  it.  Much  more  striking  instances  (textiles  at  the  beginning 
of  the  nineteenth  century,  power  at  the  beginning  of  the  twentieth,  for 
example)  could  easily  be  given.  The  group  price  for  farm  products,  far 
from  contradicting  the  principle,  really  strengthens  the  evidence  for  it. 
There  was  plenty  of  innovation,  both  in  the  production  and  in  the  trans- 
portation of  farm  products,  but  since  both  opened  European  markets 
for  American  exports,  it  is  in  perfect  accordance  with  our  point  of  view 
as  well  as  with  general  theory  that  this  group  price  did  not  fall  much  in  the 
United  States.  In  England  it  fell.  In  Germany  it  did  not,  but  this  of 
course  is  due  to  the  protection  to  agriculture  which  set  in  at  the  end  of  the 
seventies. 

The  cyclical  behavior  of  group  prices  of  the  second  class  is  conditioned 
by  the  cyclical  effect  of  the  properties  that  the  constituents  of  each  group 
have  in  common.  Monopolistic  pricing  is  an  example  for  such  properties 
and  illustrates  the  proposition  that  they  primarily  influence  the  response 
to  the  impact  of  the  active  elements  of  the  cyclical  process  (see  Chap.  X). 

But  in  group  prices  of  the  third  class  we  shall  expect  fundamental 
covariation.  Although  entrepreneurial  activity  impinges  on  particular 
spots  in  the  system,  the  effects  of  the  increase  in  producers*  expenditure 
in  prosperity  and  of  the  decrease  of  it,  in  amount  or  rate,  toward  the  end 
of  prosperity  and  in  recession,  will  make  themselves  felt  very  quickly 
all  over  the  system.  Increasing  wages,  in  particular,  are  promptly  spent 
in  prosperity  (qualifications  of  secondary,  if  yet  considerable,  importance 
will  be  noticed  at  a  later  stage)  and  under  the  influence  of  optimistic 
expectation  (greased  in  many  cases  by  ready  availability  of  consumers' 
credit)  expenditure  may  even  outrun  actual  receipts,  as  pessimistic 
expectation  may  cause  consumers'  expenditure  to  decrease  by  more  than 

1  This  should  be  emphasized.  Many  economists  are  in  the  habit  of  arguing  as  if  innova- 
tion affected  relative  prices  only  (as,  in  fact,  it  does  directly)  and  as  if,  hence,  movements  of 
the  price  level  could  have  nothing  to  do  with  it.  But  it  is  easy  to  see  that  these  movements 
cannot  be  treated  as  an  independent  component  of  individual  prices. 


480  BUSINESS  CYCLES 

the  actual  decrease  in  receipts,  as  soon  as  a  depression  gets  under  way. 
There  is,  hence,  no  theoretical  reason  to  think  that  consumers'  goods 
prices  must  always  or  necessarily  or  significantly  lag  behind  producers' 
goods'  prices.  Nor  is  there  any  reason  why  they  should  precede.  From 
our  theory  (Chap.  IV  Sec.  A)  we  should  rather  expect  almost  synchronous 
movements  in  the  same  direction  in  all  sectors  of  the  economic  organism, 
and  differences,  in  a  big  country,  to  be  more  regional  than  "stagewise." 
That  this  is  so,  is  the  most  important  result  to  be  gleaned  from  any  study 
of  group  prices  of  this  class.  The  facts  do  not  lend  much  support  to 
those  views  about  the  cyclical  mechanism  from  which  characteristic 
sequences  would  follow,  and  least  of  all  to  the  theories  about  savings 
deflecting  "purchasing  power"  from  consumers'  goods  to  producers' 
goods.1 

This  does  not  mean  that  deviations  from  synchronous  parallelism  are 
either  unimportant  or  uninteresting.  They  are  more  important  in  the 
first  two  classes  of  group  prices,  but  they  are  not  absent  in  the  third  class. 
It  would  be  most  astonishing  if  there  were  no  such  deviations,  because 
each  group  price  combines  such  different  patterns  of  action  and  response 
that  it  would  be  little  short  of  miraculous  if  the  outcome  were  equal  and 
synchronous  percentage  change  for  all.  But  the  point  is  that  we  cannot 
generalize  about  those  differences,  which  have  their  roots  in  the  peculiar- 
ities of  the  individual  commodities  entering  each  group  and  in  the  con- 
ditions of  each  cycle  and,  therefore,  teach  little  about  the  fundamental 
mechanism  of  cycles  in  general.  Thus  the  reader  will  see  from  our  chart 
that  the  producers'  goods  price  composite,  although  it  displays  sub- 
stantially the  same  rhythm  as  the  consumers'  goods  price  composite, 
shows  greater  amplitudes  and  generally  also  some  precession.2  But 

1  It  is  perhaps  worth  mentioning  that  one  of  the  many  mistaken  interpretations  of  the 
analytic  schema  presented  in  this  book  was  to  the  effect  that,  according  to  it,  prices  of  con- 
sumers' goods  should  be  the  first  to  fall  and  the  ones  to  create  "the  trouble."     This  is  a 
complete  misunderstanding  of  the  proposition  that  in  the  end  the  cyclical  process  of  evolu- 
tion means  increased  real  income  to  all  classes,  conveyed  to  them  primarily  through  falling 
price  levels. 

2  One  of  the  first  attempts  at  systematic  comparison  between  producers*  and  consumers' 
goods  prices  was  that  of  Professor  Mitchell  (see  Bureau  of  Labor  Statistics  Bulletin  178). 
The  period  between  1890  and  1918  affords  instances  for  the  above:  the  wider  fluctuations  in 
producers'  goods  prices  at  wholesale  stand  out  clearly  enough.    They  fell  sharply  to  1894 
and  less  than  consumers'  goods  prices  to  1895.     Then  they  rose  in  1896,  while  consumers' 
goods  prices  were  still  falling;  after  a  setback  they  rose  together  with  consumers'  goods 
prices,  overtaking  them  in  1898.     They  did  not  fall  in  1908  and  1904  when  consumers' 
goods  prices  did.     In  1909  consumers'  goods  prices  rose,  while  producers'  goods  prices 
were  still  falling.     In  1913  the  situation  was  similar.     But  even  as  far  as  producers'  goods 
prices  can  be  said  to  lead,  it  will  be  more  true  to  reality  to  stress  the  fact  that  the  people 
who  buy  them  have,  as  a  class,  a  wider  range  of  foresight,  rather  than  the  fact  of  their 
position  in  the  productive  process. 


THE  PRICE  LEVEL  481 

this  is  not  wholly  due  to  greater  distance  from  the  demand  of  ultimate 
consumers,  which,  though  it  actually  is  comparatively  steady,  yet 
fluctuates  quite  appreciably  in  the  course  of  cyclical  phases.  There  are 
raw  materials  that  have  only  a  short  way  to  go  before  reaching  households 
and  yet  fluctuate  as  much  as  some  others  that  are  a  long  way  off.  Com- 
parative steadiness  of  consumers'  goods  prices,  such  as  it  is,  is  to  a  large 
extent  a  result  of  a  quite  different  factor,  viz.,  that  consumers'  goods 
prices,  even  if  not  retail  prices,  include  more  branded  and  serviced 
articles  than  the  producers'  goods  group.  Hence,  we  must  not  trust  too 
much  to  mere  distance  from  consumption,  but  also  look  for  other  explana- 
tions, such  as  durability,  which  accounts  for  the  violent  fluctuations  in 
the  group  price  for  metals  and  metal  products.  Certainly  entrepreneurs' 
demand  impinges  conspicuously  there.  But  it  also  impinges  on  labor 
and  thus  on  consumers'  goods,  and  there  is  no  cogent  reason  for  assum- 
ing, a  priori,  that  the  one  effect  must  be  stronger  than  the  other. 
From  the  standpoint  of  our  model  of  the  cyclical  process  this  is  largely 
accidental.1 

While  these  considerations  tell,  as  in  fact  they  are  intended  to,  against 
theories  which  assume  systematic  relations  between  group  prices  that  do 
not  exist  at  all  or  do  not  assert  themselves  as  clearly  as  they  would  have 
to  in  order  to  prove  anything  for  these  theories,  they  do  not  tell,  nor  are 
they  intended  to  tell,  against  a  proposition  dear  to  some  "endogenous" 
or  "self -generating"  theories,  viz.,  that  as  prosperity  wears  on,  costs  of 
doing  business  increasingly  encroach  upon  profits.  We  do  not  consider 
this,  taken  by  itself,  to  be  a  satisfactory  explanation  of  the  turn  of 
prosperity  into  recession.  Nor  are  we  able  to  accept  as  relevant  all  the 
factors  that  have  been  listed  as  contributory — losses  due  to  bad  debts, 
for  example,  hardly  increase  before,  or  independently  of,  the  turn.  But 
we  do  not  deny  the  reality  of  the  mechanism,  which  indeed  forms  a  part 
of  our  model.  Many  "old"  firms,  as  pointed  out  in  Chap.  IV,  will  be 
inconvenienced  from  the  outset  by  the  rising  costs  of  labor,  materials 
(which  however  again  include  labor),  credit,  and  so  on,  and  most  "old" 
firms — all  in  fact  that  do  not  receive  more  than  their  share  of  the  expendi- 
ture induced  by  entrepreneurial  activity — will  get  into  difficulties  as  soon 
as  entrepreneurial  activity  begins  to  slacken,  even  apart  from  the  effects 
of  direct  competition  by  "new"  plants  and  from  the  difficulty,  frequently 

* *  Covariation,  tempered  by  peculiarities  of  the  industries  involved,  is  also  the  outstand- 
ing feature  of  the  movements  of  foods  and  materials.  Here,  if  anywhere,  should  we  expect 
neighborhood  of  the  sphere  of  consumption  to  assert  itself,  since  the  demand  for  food  is 
certainly  much  more  cyclically  stable  than  the  demand  for  anything  else.  It  does  assert 
itself,  but  hardly  ever  to  the  point  of  discrepancy  in  direction,  the  cases  of  which  are,  more- 
over, all  explainable  by  circumstances  that  have  nothing  to  do  with  the  cyclical  process. 
See,  for  example,  the  Sauerbeck  Food  and  Material  indices,  Journal  of  the  Royal  Statistical 
Society,  1910,  p.  316;  1929,  p.  239. 


482  BUSINESS  CYCLES 

present  in  the  case  of  finished  commodities,  of  increasing  a  price  which  is 
fixed  by  custom,  business  policy,  or  public  authority.  As  Professor 
Mitchell  has  shown,  this  will,  for  many  industries,  result  in  buying  prices 
of  materials  gaining  on  selling  prices  of  products.  This,  however,  has 
only  a  distant  connection  with  the  relative  movement  of  producers' 
and  consumers'  goods  prices. 


CHAPTER  IX 

Physical  Quantities.  Employment 


A.  Individual  and  Composite  Quantities. — As  in  the  case  of  prices, 
we  have  data  about  quantities  produced,  forwarded  to  mills,  transported, 
exported,  imported,  bonded,  subjected  to  payments  of  excise,  sold  to 
consumers  (sometimes  put  equal  to  quantities  produced  plus  imports 
minus  exports),  visibly  available  or  reflected  by  such  indicators  as 
spindles  active  or  active  spindle  hours,  furnaces  in  blast,  and  the  like. 
These  series  are  very  plentiful  for  the  postwar  period,  fairly  plentiful 
from  about  1870,  and  our  stock  of  them  for  the  more  remote  past  increases 
steadily.  Many  are  available  monthly  or  even  weekly,  and  all  may  be 
said  to  mean,  approximately  at  least,  a  definite  real  thing  that  it  requires 
no  theory  to  understand.  Qualification  is  necessary,  however.  Some 
of  the  modern  and  much  of  the  older  material  is  untrustworthy  or  at  least 
inexact,  such  as  earlier  estimates  of  crops  or  of  pig  iron  produced  or,  in 
times  and  situations  in  which  smuggling  is  a  factor  of  importance, 
figures  of  imports.  Some  data  raise  questions  of  units  of  measurement 
and  comparability. 

Changes  in  quality  or  territory  and  so  on  cast  doubt  on  the  value  of 
many  series.  Changes  in  sources  and  methods  of  compilation  introduce 
spurious  breaks  and  fluctuations.  The  theorist's  questions — what  is  a 
commodity,1  a  factor  of  production,  a  country? — acquire  ominous  sig- 
nificance for  the  most  practical  purposes.  In  such  cases  as  that  of 
building  permits,  the  meaning  itself  becomes  doubtful.  We  not  only 
open  up  obvious  sources  of  serious  error,  but  also  obscure  essential 
features  of  cyclical  movements,  if  from  figures  measuring  one  stage  of  the 
career  of,  for  example,  a  raw  material,  we  draw  conclusions  about  another 
stage — from  iron,  conclusions  about  industrial  equipment;  from  wheat, 
conclusions  about  bread;  or  from  imports  or  exports,  conclusions  about 
production.  Finally,  the  changing  significance  and  efficiency  of  indi- 

1  If  we  call  motorcars  a  commodity,  we  are  immediately,  as  in  the  case  of  prices,  faced 
by  an  index  problem.  If  we  restrict  the  concept  to  model  X  of  the  firm  Yt  the  material 
becomes  unmanageable.  The  factor  labor  brings  out  the  difficulty  best  of  all.  A  country 
like  the  United  States,  or  even  France,  splits  into  sectors  much  more  different  in  character 
than  Venezuela  is  from  Colombia. 

483 


484  BUSINESS  CYCLES 

vidual  commodities  in  consumers'  budgets  and  in  the  pattern  of  pro- 
duction, itself  a  most  important  feature  of  the  results  of  the  cyclical 
process,1  invites  erroneous  interpretations  if  the  series  is  not  studied  in 
relation  to  the  history  of  its  industry  and  technology,  which  alone  gives 
the  key  to  its  meaning.  Another  research  program  unfolds  itself,  quite 
beyond  the  means  of  the  individual  worker. 

If  we  form  composites  from  groups  of  related  commodities,  such  as 
foods,  furniture,  equipment,  textiles,  and  the  like,  we  meet,  only  with 
more  inescapable  clearness,  a  problem  that  differs  but  in  degree  from 
the  problem  which  is  implied  in  speaking  of  a  quantity  of  coal  in  general 
or  coffee  in  general.  A  heap  of  woolen  and  cotton  fabrics  is  possibly 
less  wanting  in  meaning  than  a  heap  of  iron  and  strawberries  would  be. 
In  the  most  favorable  cases,  notably  in  cases  of  complementarity  and 
rivalry  in  fixed  proportions,  but  also  beyond  these,  even  exact  theoretical 
meaning  may  be  attributed  to  the  composite.  We  shall  not,  however, 
enter  into  these  problems  but,  simply  relying  on  the  common  sense  of  the 
thing,  confine  ourselves  to  the  remark  that  such  composites  derive  addi- 
tional justification  if  their  constituents  are,  owing  to  their  place  in  the 
economic  organism,  all  exposed  to  similar  external  and  internal  influences. 
As  with  group  prices,  it  is  necessary  to  bear  in  mind  that  composites  may 
seriously  obscure  what  precisely  is  the  essential  fact  about  the  cyclical 
process  of  economic  evolution. 

But  the  concept  of  total  output  lacks  similar  meaning  or  justification. 
The  fact  that  it  has  gained  citizenship  and  is  "recognized"  is  no  comfort, 
for  this  recognition  has  been  extended  to  it  quite  uncritically.  The  prob- 
lem is  much  more  doubtful  than  the  corresponding  one  in  the  case  of 
prices.  There,  we  have  at  least  been  able  to  discover  and  define  an 
economic  magnitude  that  supplied  the  meaning  of  an  index  of  the  price 
level.  Here,  no  such  thing  actually  exists.  However  useful  for  many 
purposes,  total  output  is  a  figment  which,  unlike  the  price  level,  would 
not  as  such  exist  at  all,  were  there  no  statisticians  to  create  it.  We  seem 
indeed  to  be  faced  by  a  meaningless  heap. 

Three  ways  are  open  to  us  in  order  to  overcome  this  difficulty,  although 
none  of  them  can  be  considered  as  entirely  satisfactory.  The  first  has  the 

1  In  particular,  products  require  as  time  goes  on,  less  and  less  of  raw  materials,  such 
as  coal,  steel,  and  sugar  beet,  per  unit.  For  many  purposes,  series  should  be  corrected  for 
this.  It  certainly  vitiates  the  implications  of  rates  of  increase  in  the  production  of  those 
materials.  The  use  of  scrap  is  an  instance  of  difficulties  of  another  type.  Improving 
quality,  still  another  striking  feature  of  economic  evolution,  works  the  same  way.  It  also 
largely  escapes  us.  We  gather  from  historical  indications  that  even  the  oldest  and  most 
ordinary  articles  of  consumption,  such  as  meat  and  wine,  are  quite  different  now  from  what 
they  were  even  100  years  ago.  But  in  most  cases  we  have  no  means  of  measuring  the 
change.  This  really  casts  doubt  on  the  possibility  and  meaning  of  any  statement  that 
turns  on  any  but  the  most  outstanding  features  of  our  graphs. 


PHYSICAL  QUANTITIES.    EMPLOYMENT  485 

merit  of  simplicity.  It  was  mentioned  in  the  first  chapter  and  consists  in 
fastening  upon  some  series  the  items  of  which  measure  something  that 
may  be  taken  to  indicate  the  variations  in  productive  activity — cyclical 
variations  in  particular.  All  quantity  series  have  some  "systematic" 
significance,  but  those  we  have  decided  to  call  systematic  have  much  more 
of  it  than  others.  Their  limitations  are,  however,  serious.  Employ- 
ment, even  if  statistics  for  our  period  were  better  than  they  are,  would  be 
seriously  misleading  for  periods  longer  than,  say,  a  Juglar,  often  for  even 
much  shorter  periods:  in  good  theory,  it  can  never  be  assumed  to  be 
proportional  to  output  (see  Sec.  D).  Tonnage  cleared,  not  valueless  for 
Great  Britain,  is  open  to  objections,  not  only  because  of  the  changing 
significance  of  the  unit  and  the  changing  importance  of  international  trade, 
but  because  it  puts  a  ton  of  coal  equal  to  a  ton  of  gramophone  records. 
Similar  difficulties  surround  the  use  of  figures  of  freight  carried  by  rail- 
roads and,  for  example,  horsepower  installed.  Pig-iron  consumption, 
though  of  course  not  above  criticism,  does  astonishingly  well  for  prewar 
times,  if  tested  in  the  light  of  our  historical  knowledge  of  cyclical  fluctua- 
tions.1 We  will  therefore  immediately  present  the  series  and  also  the 
series  of  rates  of  change  of  pig-iron  consumption,  Charts  XII  and  XIII. 

The  trends  in  Chart  XII,  of  course,  are  the  trends  of  pig-iron  consump- 
tion and  of  nothing  else.  They  do  not  give  the  trends  of  total  output, 
since  the  equipment  into  which  pig  iron  enters  increased  faster  than  other 
elements  of  total  output,  nor  do  they  give  even  the  trend  of  total  "real" 
investment.  Moreover,  pig  iron  cannot  be  trusted  to  give  the  cyclical 
variation  of  total  output,  since  cyclical  variations  in  consumers'  goods' 
production  are  not  proportional  to,  or  synchronous  with,  cyclical  varia- 
tions in  the  production  of  equipment.  But  the  fluctuations  in  the  produc- 
tion of  equipment  (steel  did  not  in  our  period  enter  so  largely  into 
consumers'  goods  as  it  does  now)  in  all  three  cycles  it  gives  very  well.2 
This  is  particularly  true  of  the  Juglar,  which  it  is  hardly  possible  to  mark 
off  so  neatly  in  any  other  series — the  reader  should  mark  it  off  here  and 
use  the  result  for  reference.  The  three  curves  in  a  sense  interpret 
each  other.  The  American  curve,  otherwise  the  most  satisfactory  one 
and  that  which  shows  the  firmest  step,  presents  a  problem  for  1875  to  1885 

1  Pig  iron  held  its  own  fairly  well  until  about  20  years  ago,  though  the  process  of  ousting 
it  had  set  in  by  the  sixties.    Later,  castings  were  quickly  going  out  and  weldings  became 
of  much  Jess  weight.    Lighter  alloys  also  came  in,  and  other  metals  gained  relatively, 
particularly  aluminum.    The  increasing  use  of  scrap  has  been  mentioned.     On  the  other 
hand,  iron  and  steel  gained  ground  at  the  expense  of  timber  and  other  materials.     As 
has  been  mentioned,  consumption  of  pig  iron  has  been  used  as  the  backbone  of  the  analysis 
of  cycles  by  Professor  Spiethoff. 

2  There  is  some  interest  in  noting  that  England's  pig-iron  consumption  hardly  took  part 
at  all  in  the  Kondratieff  prosperity  after  1897 — another  indication  of  the  fact  that  the 
entrepreneurial  impetus  in  English  industry  began  to  slacken  about  that  time. 


BUSINESS  CYCLES 


V 


\ 


fft 

58  L 

IS  « 


I     I      I      I       I         I 


*-> 


ex 

M 


I 


PHYSICAL  QUANTITIES.    EMPLOYMENT 


487 


that  we  know  how  to  solve  as  soon  as  we  glance  at  the  English  a^nd 
German  lines,  the  behavior  of  which  is  in  turn  cleared  up  by  the  American 
line  in  the  segment  1885  to  1892.  The  differences  between  the  three  lines 
in  average  amplitudes  and  general  character  are  very  revealing  of  the 
pulse  of  economic  evolution  in  the  three  countries.  The  Kitchins,  so  the 
writer  thinks,  show  with  sufficient  clearness  on  Chart  XIII.  No  doubt 
some  readers  will  not  agree.  To  probable  objections  it  can  only  be  replied 
that  it  seems  a  mistake  to  refuse  to  consider  as  a  distinct  class  fluctuations 
which  are  so  similar  in  nature  and  so  clearly  marked,  only  because  they  do 


Per 
Cent 

+20 


UNITED  STATES 


UNITED  KINGDOM 
A  / 


A 


-20 
+20 

0 
-20 

+20 
0 

-20 

1855       1860        1865        1870        1875       1880        1885        1890       1895        1900        1905        1910 

CHART  XIII. — Rate  of  percentage  change  of  pig  iron  consumption  (see  Appendix,  p.  1056). 

not  display  a  regularity  that  simply  is  not  in  the  phenomenon.  Chart 
XIV  is  presented  in  order  to  indicate  how  far  other  series  of  more  than 
average  systematic  significance  could  render  the  same  service. 

We  may,  in  the  second  place,  define  variations  in  the  "physical 
volume"  of  any  composite  to  mean  variations  in  its  dollar  volume  cor- 
rected for  changes  in  price  level.  The  result,  it  should  be  noted,  is  still  a 
value,  and  not  a  true  physical  quantity,  as  we  might  expect  according  to 
those  definitions  of  price  level  which  make  it  a  sort  of  average  price. 
Our  level  is  not  a  price  but  a  pure  number.  Hence  "deflated  dollar  vol- 
ume" is  simply  a  value  figure  from  the  variations  of  which  the  effects  of 
the  variations  of  the  monetary  parameter  have  been  removed.  However, 


V 


..A 


GERMANY 


A 


488 


BUSINESS  CYCLES 


COAL  PRODUCTION 


/ 


COTTON   »RODUCTIONy 


/ 


V 


RAILWAY  FREIGHT  TON-MILES 


f 


/BUILDING  PERM  |rS 


COTTON  CONSUMPTION 


Vw' 


1840      45        I860        55        1860       65        1870       75         1880       85        1890       95        1900       05       1910    13 
CHART  XIV.— United  States  (see  Appendix,  p.  1056). 


PHYSICAL  QUANTITIES.    EMPLOYMENT  489 

this  has  definite  meaning,  and  these  corrected  figures  may  for  some  pur- 
poses be  treated  as  if  they  were  physical  quantities.  It  should  be  observed 
that,  in  order  to  effect  this,  it  is  necessary  to  deflate  by  an  index  which 
approximately  presents  our  level  concept,  and  not  by  any  other — for 
example,  an  index  specially  constructed  to  include  only  the  prices  of  those 
commodities  that  enter  into  the  given  composite,  with  weights  that  cor- 
respond to  the  relative  importance  of  those  commodities  in  the  composite. 
We  are  not  concerned  with  the  question  whether  the  latter  proceeding  has 
any  meaning  of  its  own.  All  that  matters  is  that  this  meaning  is  a  dif- 
ferent one. 

The  money  value  of  Total  Output  (unfortunately  along  with  other 
items)  is  reflected  in  Outside  Clearings,  the  value  of  Total  Output  of 
Consumers'  Goods,  in  the  Sum  Total  of  Private  Incomes  minus  savings 
and  taxes  paid  out  of  income.  Where  we  have  these  two  figures,  or  at  any 
rate  one  of  them,  we  might  hence  try  to  solve  the  problem  in  hand 
by  deflating.  The  reasons  why  we  do  not  primarily  rely  on  this  possibility 
will  be  apparent  later  when  we  discuss  the  Clearings  and  Income  series. 
Insuperable  theoretical  scruple  is  not  one  of  them.  But  although  we  make 
bold  to  use  as  a  level  series  what  we  know  to  be  something  else,  and 
although  we  may  be  equally  bold  in  the  case  of  Clearings  and  Incomes 
taken  by  themselves,  we  hesitate  to  cumulate  errors  by  combining  the 
two,  even  apart  from  various  other  objections  on  the  score  of  statistical 
method.1 

A  third  method  of  arriving  at  a  single  figure  indicative  of  variations 
of  total  physical  quantities  of  commodities  produced  or  consumed  (services 
and  voluntary  leisure  ought,  strictly  speaking,  to  be  included),  consists  in 
constructing  an  index  from  individual  quantity  series.  Everybody 
recognizes  that  quantities  produced,  consumed,  and  in  stock,  ought  to  be 
separately  combined  and  that  these  separate  indices  should  be  confronted, 
but  everybody  puts  up  with  a  combination  of  everything  that  is  to  be  had. 
The  leading  contributors  all  have  a  preference  for  weighting  individual 
relatives,  or  relatives  adjusted  for  some  trend  and  for  seasonal  variations, 
with  the  American  Value  Added  by  Manufacture  or  the  English  Net 
(Value)  Product,  where  products  of  different  stages  of  a  productive  process 
are  included.  But,  in  cases  in  which  these  values  are  not  available,  some 
fall  back  upon  other  criteria  of  relative  importance,  such  as  workmen 
employed,  pay  rolls,  horsepower  installed,  and  so  on — none  of  which  they 

1  Mr.  Snyder's  "A  New  Clearings  Index  of  Business  for  Fifty  Years."  Journal  of  the 
American  Statistical  Association,  1924  p.  829,  does  precisely  this,  except  that  he  deflates  by 
his  General  Price  Level,  which,  for  the  case  in  hand,  is  an  index  specially  constructed  to 
correspond  to  the  various  components  of  the  Clearings  figure.  The  above  argument  is  not 
intended  to  imply  adverse  criticism  of  this  interesting  experiment. 


490  BUSINESS  CYCLES 

would  defend  on  theoretical  grounds.1  The  formal  theory  of  such  an 
index  is  easy  to  derive.  All  we  need  to  do  is  to  change  the  Laspeyres 
formula  for  the  price  level  (see  Chap.  VII,  Sec.  B)LP  =  2piq0/2p0q0  into 


the  obviously  equivalent  expression  Lp  =  —  —^  -  ,  where  the  values 

^PoQo 

p0<lo  are,  in  the  numerator,  now  used  as  weights  of  price  relatives,  and  to 

* 

form   the  analogous   expression  for  quantities,  viz.    Lq 

which  means  quantity  relatives  weighted  by  values,  though  not  by  added 
values.  But  putting,  as  we  did  before  in  the  case  of  prices,  q%  =  q0  +  dqot 
and  dropping  subscripts,  we  get 

1  There  are  available  for  the  postwar  time  and  particularly  for  this  country  indices  of 
this  kind  which  excel  in  careful  construction  and  technical  perfection.  Given  the  limitations 
imposed  by  the  material  and  granted  the  principles  upon  which  their  construction  proceeds, 
they  probably  do  as  much  as  can  fairly  be  asked  in  all  such  matters  as  assembling  and 
criticizing  the  material,  and  correcting  for  seasonal  variations  and  differences  in  working 
days.  The  pioneer  work  of  Mr.  E.  E.  Day  and  the  Harvard  Society,  its  continuation  and 
amplification  by  the  Federal  Reserve  Board,  Prof.  Warren  Persons'  and  Mr.  Snyder's 
contributions,  the  indices  of  the  Standard  Statistics  Company  and  of  Mr.  Leong,  the 
English  index  of  Mr.  Rowe  and  the  London  and  Cambridge  Economic  service,  and  the 
German  indices  of  the  Institut  filr  Konjunkturforschung  must  be  specially  and  gratefully 
mentioned.  In  particular,  the  distinction,  fundamentally  important  as  we  shall  presently 
see,  between  equipment  goods,  other  producers',  and  consumers'  goods,  has  been  fully 
attended  to.  For  the  prewar  time  we  are  less  well  off,  of  course.  There  is,  again,  Mr. 
E.  E.  Day's  work  and  the  work  of  Professor  Warren  Persons.  Earlier  pioneers  ought  to  be 
mentioned,  notably  E.  Leonard:  Index  of  Changes  in  Extractive  Industries  (Publication  of 
the  American  Statistical  Society  for  September  1918)  and,  although  their  investigation 
bore  on  trade  rather  than  production,  Professors  E.  W.  A.  Kemmerer  and  Irving  Fisher. 
The  subject  is  also  indebted  to  Professor  W.  J.  King,  Mr.  W.  W.  Stewart  (see  the  latter's 
Index  of  Production,  American  Economic  Review  for  March  1921)  and  Mr.  W.  Thomas. 
Mr.  Snyder  extended  his  researches  back  within  his  work  on  his  index  of  trade.  Professors 
Warren  and  Pearson  have  made  an  extremely  valuable  contribution  as  to  basic  and  agricul- 
tural production.  The  Index  of  General  Business  of  the  American  Telephone  and  Tele- 
graph Company  has,  in  1922,  been  reduced  to  an  exclusively  physical  index.  Colonel 
Ayres'  index  contains  also  prices  and  values  from  1790  to  1855,  but  is  exclusively  physical 
from  1855  to  1901,  with,  however,  such  weighting  as  would  make  it  lean  on  the  Day-Thomas 
index  used  by  him  for  1901  to  1919.  For  England  there  were  only  various  indicators  and 
individual  bits  of  information  (for  the  early  part  of  our  period,  largely  contained  in  Porter 
and  Baxter),  but  we  may  now  use  the  index  compiled  by  Mr.  Hoffmann  of  the  Kiel  Institut 
ftir  Seeverkehr  und  Weltwirtschaft,  which,  however,  had  to  work  with  material  not  admit- 
ting of  any  rational  system  of  weighting.  It  goes  back  to  1713  but  we  use  it  only  from  1785, 
because  before  that  date  component  series  are  too  few.  Mr.  Snyder's  index  of  English 
production  should  also  be  mentioned.  For  Germany,  we  use  the  Index  of  the  Institut  flir 
Konjunkturforschung,  compiled  by  Mr.  Wagenfiihr,  who  has  had  to  substitute,  in  some 
cases,  for  figures  of  production  or  consumption,  figures  of  net  import,  railway  transporta- 
tion, or  even  money  value. 


PHYSICAL  QUANTITIES.    EMPLOYMENT  491 


that  is  to  say,  an  index  of  that  part  of  the  actual  relative  variation  in 
expenditure,  (E  +  dE)/E,  which  as  we  have  put  it  before,  is  balanced  as 
to  its  effects  on  the  price  level  by  changes  in  physical  quantities  and  by 
which  expenditure  would  have  had  to  change  in  order  to  keep  the  price 
level  constant. 

Hence  we  have  again  a  value  figure  from  which  the  effects  of  changes 
in  the  monetary  parameter  have  been  removed.  The  method  is  thus 
seen  to  be  but  a  variant  of  the  method  of  deflating  values.  It  aims  at 
what  is  essentially  the  same  thing.  However,  it  not  only  avoids  the 
difficulties  that  arise  from  the  nature  of  the  clearings  figure  and  the 
cumulation  of  errors  incident  to  the  process  of  deflating,  but  is  theoretically 
superior  to  the  latter,  in  that  it  follows  logically,  and  derives  its  meaning, 
from  the  equation  which  embodies  the  theory  of  the  price  level  and  can, 
owing  to  this  fact,  never  give  absurd  results.1  If  we  recall  how  in  the 
fourth  chapter  we  defined  the  effects  of  innovation  on  the  depth  and 
breadth  of  the  stream  of  commodities,  we  see  immediately  in  what  sense 
such  an  index  implements  our  propositions  about  variation  of  output  in 
the  course  of  business  cycles.  It  does  not  measure  physical  output  in  the 
literal  sense,  but  it  docs  measure  physical  output  as  transformed  by  the 
introduction  of  an  economic  dimension.  An  argument  would  be  in  order 
here  for  taking  account  of  the  structure  of  the  world  of  commodities  by 
the  choice  of  commodities  to  be  included.  But  since  it  would,  mutatis 
mutandis,  only  repeat  what  has  been  said  on  this  question  in  our  discus- 
sion of  the  level  index,  we  will  not  stay  to  develop  it. 

B.  The  Analysis  of  the  Trend  in  Total  Industrial  Output.  —  Quantities 
produced  or  consumed  are,  unlike  prices,  rates  per  time  element.  Their 
variations  in  the  course  of  cycles  are  part  of  the  primary  as  well  as  of 
the  secondary  phenomena  and,  within  both  these  categories,  consequential 
with  respect  to  some,  causal  with  respect  to  others  —  the  most  important 
approach  to  a  primary  causal  role  being  the  relation  of  actual  or  expected 
release  of  new  products  to  the  turn  from  prosperity  into  recession.  An 
index  of  Total  Output  makes  a  synthetic,  systematic,  and  cyclical  trend 
series. 

To  verify  expectation  as  to  the  presence  of  a  trend,  it  is  only  necessary 
to  look  at  the  quantity  lines  in  charts  V,  VI,  and  VII.  Moreover,  Chart 

1  It  should  be  noticed  that  the  Harvard  Index  of  Physical  Volume,  which  is  in  essence 
such  a  corrected  value,  in  the  short  run  goes  very  well  with  Freight  Cars  Loaded,  if  both 
are  adjusted  for  business  days  per  month  and  for  seasonal  variations,  and  treated  according 
to  the  Harvard  method.  See  W.  M.  Persons,  Car  Loadings  as  an  Index  of  Trade  Volumes, 
Review  of  Economic  Statistics  for  October  1926.  The  same  is  true  for  Germany,  see  W. 
Teubert,  Der  Gliterverkehr,  Sonderheft  5  of  Vierteljahrahefte  zur  Konjunkturforschung,  1928. 


492 


BUSINESS  CYCLES 


XV  presents  an  international  comparison  that  is  of  some  interest, 
although,  owing  to  differences  in  materials  and  methods  it  must  not  be 
relied  on  too  implicitly.  Chart  XVI  gives  the  same  series  as  transformed 
by  the  operation  of  empirical  differentiation  and,  by  the  absence  of  any 
marked  trend,  teaches  us  something  about  the  character  of  the  trend 
present  in  the  original  figures.  Of  course,  this  does  not  mean  that  with 
other  data  the  same  result  would  emerge,  or  that  any  trust  can  be  placed 


1780   1790    1800    1810     1820     1830     1840     1850     1860     1870     1880     1890    1900     1910 
CHART  XV. — Industrial  production  (see  Appendix,  p.  1056). 

in  the  gradient  of  the  particular  logarithmic  straight  line  which  an  investi- 
gator's material  may  yield.  Some  authors  who  speak  of  a  "compound 
interest  law  of  growth"  seem,  like  the  queen  in  the  play,  to  protest  too 
much.  Still  more  treacherous  and  pregnant  with  danger  of  speculative 

temerity  may  be  the  application  of  Verhulst's  formula,  y  —  ,  _t  ,  ,  > 

which  was  intended  (1888)  to  represent  certain  features  of  organic  or  of 
similar  types  of  growth.1  Even  perfect  fit  in  the  least  square  sense  would 

1  Compare,  for  instance,  Wittstein's  formula  in  Gesetz  der  menschlichen  Sterblichkeit, 
1883.  Verhulst's  formula  is  used  (slightly  generalized)  by  A.  J.  Lotka.  The  form  6f  ,  • 
is  known  in  this  country  as  the  Pearl-Reed  curve. 


PHYSICAL  QUANTITIES.    EMPLOYMENT 


493 


y- 


iu-u 


I         11 


-  £ 


8  I 

si 

«•» 

a  & 


"8 

CW     a, 

if 

«g 

i 

8  3 


^        ol^ 


494  BUSINESS  CYCLES 

not  prove  anything.  We  are,  however,  on  somewhat  safer  ground  when 
applying  such  expressions  to  the  behavior  in  time  of  quantities  of  indi- 
vidual commodities. 

The  broad  fact  of  great  steadiness  in  long-time  increase  nevertheless 
remains,  both  in  the  sense  of  rough  constancy  of  the  gradient  of  the  trend 
and  in  the  sense  of  what,  merely  by  way  of  formulating  a  visual  impression, 
we  may  term  the  general  dominance  of  trend  over  fluctuations.  As  a  rule, 
even  strong  depressions  such  as  the  English  one  of  1825  could  hardly  be 
identified  as  such  from  the  output  graph  alone.  In  no  country  does  1873 
look  very  catastrophic.  In  America,  1884  produced  almost  no  fall  at  all. 
The  crisis  of  the  early  nineties  shows,  for  Germany,  by  only  an  incon- 
siderable dent.  In  the  long  English  series  it  happens  only  twice  that 
absolute  fall  outlasts  two  years.  In  the  case  of  Germany,  this  occurred 
only  in  1868,  1869,  and  1870;  in  America  also  but  once.  One  of  the  most 
important  tests  of  an  analytic  model  of  the  cyclical  process  of  economic 
evolution  is  whether  or  not  it  enables  us  to  understand  that  steadiness 
in  both  senses  (such  as  it  is). 

What  we  see,  is,  of  course,  only  a  descriptive  trend.  In  the  same  sense 
as  in  the  case  of  the  price  level  we  interpret  it  as  a  result  trend,  blurred  and 
deflected  by  outside  disturbances  which,  particularly  if  a  trend  line  be 
fitted  by  least  squares,  may  acquire  lasting  influence,  even  if  passing  by 
nature.  And  there  is  also  the  effect  of  growth.  As  has  been  pointed 
out  before,  no  satisfactory  method  is  known  to  the  writer  of  eliminating 
these  influences,  and  what  follows  must  be  read  with  due  regard  to  the 
qualifications  that  this  disability  implies.  The  result  trend  itself  must 
be  explained  from  the  behavior  of  output  in  the  cycles  that  generate  it. 
Before  taking  up  this  subject,  we  will  digress,  in  order  to  comment  on 
some  questions  which,  though  they  do  not  strictly  lie  on  our  path,  yet 
come  too  near  to  be  passed  unnoticed. 

It  is  reasonable  to  believe  that  if  our  analysis  enabled  us  to  isolate 
and  measure  the  result  trend  in  output  plus  the  effect  of  growth,  instead  of 
merely  teaching  us  to  recognize  their  presence  in  our  series,  the  output 
curves  Would  turn  out  to  be  steeper  than  the  descriptive  trend,  because 
practically  all  the  other  factors  which  influence  the  latter  are  of  the  nature 
of  injuries  inflicted  upon  the  economic  organism.  This  way  of  expressing 
a  patent  historical  fact  should  not  be  understood  to  imply  adverse  judg- 
ment about  what,  from  our  standpoint  and  for  our  present  purpose,  would 
have  to  be  included  among  injuries.  Many  measures  of  social  better- 
ment that  most  people  heartily  approve  of  and  many  nationalistic  policies 
that  command  fervent  allegiance  come  within  the  meaning  of  that  term, 
simply  because  they  prevent  the  economic  machine  from  working  accord- 
ing to  its  design,  but  nothing  is  further  from  the  writer's  mind  than  to 
hold  that  it  "should"  be  allowed  to  do  so,  or  that  everybody  would  be 


PHYSICAL  QUANTITIES.     EMPLOYMENT  495 

happier  if  it  were,  or  that  abundance  is  the  criterion  by  which  to  define 
welfare  or  to  judge  a  civilization.  Nevertheless,  the  descriptive  trend  of 
output  is  both  in  itself  and  because  of  the  idea,  however  vague,  which  it 
gives  us  about  what  the  combined  effects  of  evolution  and  growth  would 
be  in  the  absence  of  other  factors,  for  a  purely  economic  analysis,  a  fact 
of  importance.  In  a  sense  it  is  the  most  important  fact  about  the  eco- 
nomics of  capitalist  society  as  distinguished  from  its  culture  and  the  type 
of  man  it  creates.  In  beholding  it  we  have  before  us  a  measure  of  the 
actual  economic  result  of  humanity's  great  experiment  with  private  busi- 
ness and  the  acquisitive  principle. 

Whether  the  gradient  we  observe  is  only,  as  we  implied  above,  what  is 
left  of  an  even  greater  possibility  or  whether  the  injuries  referred  to  are  the 
outcome  of  the  capitalist  process  itself  and,  hence,  to  be  recorded  against 
it  and  inseparable  from  it,  need  not  detain  us  here  i1  this  is  a  question  that 
cannot  be  dealt  with  without  entering  into  the  theory  of  the  nature  and 
behavior  of  social  classes.  Nor  does  it  concern  us  how  capitalist  perform- 
ance compares  with  the  performance  we  could  expect,  or  could  have 
expected  at  any  time,  from  alternative  institutional  arrangements. 
Theories  that  seem  convincing  and  are  sympathetic  to  some  have  been 
offered,  in  order  to  make  out  a  case  for  the  proposition  that  the  economic 
engine  characterized  by  private  enterprise  necessarily  turns  out  the 
maximum  of  output  which  it  is,  or  was,  possible  to  produce  within  the 
framework  of  natural,  technological  and  political  data.  Other  theories 
that  seem  convincing  and  are  sympathetic  to  others  have  been  offered 
in  support  of  the  proposition  that  it  is  of  the  essence  of  the  system  of 
private  enterprise  to  produce  less  than  could  be  produced  within  the  same 
circumstances.  There  is  little  to  choose,  as  regards  looseness  of  logic  and 
carelessness  about  facts,  between  the  two  types  of  argument.  We  will 
confine  ourselves  to  noticing  the  following  points. 

First,  from  the  indubitable  fact  that  the  capitalist  arrangement  tends 
to  draw  the  best  brains  into  business  pursuits  and  to  spur  them  to  do  their 
utmost,  it  does  not  follow  that  these  efforts  must  necessarily  issue  in  maxi- 
mum production,  although  that  fact  remains  relevant  to  the  argument. 

Second,  for  perfectly  competitive  conditions  it  can  be  proved,  with 
but  unimportant  exceptions,  that  states  of  perfect  equilibrium  will 
in  fact  be  characterized  by  maximum  output.  This  proposition  is 
not  made  entirely  valueless  or  tautological  by  all  that  has  to  be  assumed 
1  We  met  that  question  before  in  our  discussion  of  the  concept  of  external  factors  and 
repeatedly  in  our  historical  sketch.  We  take  here  the  same  standpoint  as  we  did  there: 
there  is  obvious  sense,  we  admit,  in  saying  that  if  a  man  takes  to  drink  because  of  ill-health, 
the  effects  of  the  alcohol  are  to  be  included  in  a  full  description  of  the  working  of  his  organ- 
ism. But  they  are  still  due  to  a  distinguishable  factor  which  for  many  purposes  should  be 
dealt  with  separately.  We  shall  not  leave  this  standpoint  except  for  a  short  passage  in 
Chap.  XIV. 


496  BUSINESS  CYCLES 

in  the  process  of  proving  it.  Refinements  of  analysis  often  render  what 
amounts  to  a  disservice  in  such  matters  by  unduly  emphasizing  qualifica- 
tions of  little  practical  significance.  But  the  analogous  proposition 
can  similarly  be  proved  for  the  case  of  a  socialist  organization. 

Third,  any  argument  to  the  contrary  that  runs  on  the  lines  of  economic 
principle  at  all  must  mainly  rest  on  imperfections  of  either  competition  or 
equilibrium.  These,  however,  yield  so  strong  and  so  universally  recog- 
nized a  case  against  the  pretended  efficiency  of  the  capitalist  machine  that 
it  becomes  necessary  to  recall,  on  the  one  hand,  that  a  system  in  which 
imperfect  competition  prevails  will,  contrary  to  established  opinion, 
produce  in  very  many,  perhaps  in  most  cases,  results  similar  to  those  which 
could  be  expected  from  perfect  competition  and,  on  the  other  hand,  that 
even  if  a  system  consistently  turned  out  less  than  its  optimum  quantity, 
this  would  not  in  itself  constitute  disproof  of  optimal  performance  over 
time. 1  Whoever  wishes  to  hold,  either  that  imperfect  competition  entirely 
inverts  the  working  of  the  system,  or  that  failure  to  produce  instantaneous 
maxima  spells  failure  to  evolve  along  a  path  of  maximum  production 
in  the  sense  alluded  to,  will  therefore  have  to  embark  upon  a  much  more 
arduous  analysis  than  is  implied  either  in  reasoning  from  a  few  assump- 
tions or  in  analyzing  individual  facts  without  relating  them  to  the  whole  of 
the  economic  process.  He  will  be  on  much  safer  ground,  however,  if  he 
starts  his  attack  from  other  than  purely  economic  standpoints. 

Fourth,  whoever  wishes  to  hold  that  capitalist  performance  in  fact 
lives  up  to  the  possibilities  which  the  capitalist  system  itself  creates,  and 
that  either  the  descriptive  trend,  or  what  one  may  conceive  the  true 
result  trend  plus  growth  to  be,  really  represents  the  possible  maximum 
output  over  time,  is  of  course  under  the  same  necessity.  Such  a  venture, 
to  which  theory  can  only  contribute  a  technique  of  inference,  is  in  the 
present  state  of  our  factual  knowledge  probably  beyond  the  means  of  the 
individual  worker.2 

1  For  any  dynamic  system  (economic  or  other)  may  work  in  such  a  manner  that  it 
produces  its  optimum  (with  reference  to  any  criterion)  over  time,  only  at  the  price  of  never 
reaching  it  at  any  point  of  time.  This  is  not  more  paradoxical  than,  for  instance,  the  fact 
that  producing  at  minimum  cost  over  time  may,  in  the  case  of  lumpy  factors,  imply  never 
producing  at  minimum  cost  on  any  of  the  short-time  cost  curves  because  another  "method" 
becomes  more  advantageous  before  that  point  is  reached.  The  proposition  may  further  be 
illustrated  by  that  kind  of  excess  capacity  which  is  an  unavoidable  incident  of  quick 
"progress."  The  first  of  the  two  propositions  contained  in  the  above  sentence  may  be 
illustrated  by  that  case  of  monopolistic  competition  in  which  there  is  a  great  number  of 
actual  and  possible  varieties  of  a  commodity  that  are  highly  substitutable  for  each  other. 
Part  of  what  has  been  said  on  imperfect  competition  in  the  second  chapter  will  help  to 
establish  the  point. 

*  Many  engineers  and  efficiency  experts,  particularly  of  the  technocratic  type,  are  not 
of  this  opinion  and  find  it  extremely  easy  to  collect  overwhelming  evidence  of  underutiliza- 
tion  of  resources.  But  this  only  proves  that  they  have  not  even  seen  the  problem. 


PHYSICAL  QUANTITIES.     EMPLOYMENT  497 

But  whether  or  not  the  descriptive  trend  traces  out  a  path  of  maximum 
production — be  this  maximum  relative  to  capitalist  conditions  or  to  more 
general  ones — the  fact  remains  in  any  case  that  such  performance  as  that 
trend  embodies  is  not  only  historically  associated  with  that  concentration 
of  human  efforts  on  private  economic  ends,  with — perfectly  or  imperfectly 
— competitive  private  enterprise,  with  intensive  saving,  in  short  with 
"profit  economy,"  but  also  either  wholly  or  to  a  considerable  extent 
historically  dependent  upon  it :  so  much  is  clear  from  our  analysis  and  so 
much  would — witness  the  Communist  Manifesto-^have  been  admitted  by 
Marx  himself,  though  not  by  many  of  his  followers.  Since  extrapolation 
of  the  descriptive  trend  yields  the  result  that  the  profit  economy  would,  if 
allowed  to  work  on,  do  away,  in  a  not  distant  future,1  with  anything  that 
can  according  to  present  standards  be  called  poverty,  it  is  of  some  interest 
to  inquire  whether  there  is  any  warrant  for  such  an  extrapolation.  This 
raises  the  question  of  the  phenomenon  of  retardation  in  rates  of  increase, 
which  some  students  believe  our  physical  series  display.2 

Within  the  range  of  problems  we  are  envisaging  at  the  moment,  the 
question  carries  meaning  only  for  total  output  or  total  output  of  con- 
sumers' goods.  For  it  is  obvious  that  even  if  we  disregard  the  phe- 
nomenon, fundamental  to  economic  evolution,  of  displacement  of  old 
commodities  by  new  ones,  no  industry  can  go  on  expanding  output  at  the 
rate  of  its  innovation  stage.  Each  reaches  maturity  in  the  sense  that  it 
finds  its  place  in  the  economic  organism  and  the  amount  of  output  beyond 
which  it  cannot  profitably  go,  unless  that  amount  be  increased  by  some 
further  innovation  within  it  or  in  some  "complementary"  industry  and 
by  the  general  effects  of  increasing  wealth  (which,  however,  may  also  be 
negative)  and  of  Growth.  But  it  is  equally  obvious  that  by  itself  neither 
absolute  rate  of  increase  nor  rate  of  increase  per  head  of  population 
corrected  for  age  distribution,  will  give  us  the  information  required,  for, 
to  exemplify  by  a  strong  case,  absolute  decline  in  population  may  so 
reduce  the  wants  of  the  community  that  even  strong  and  persistent 
progress  in  efficiency  of  the  productive  engine  could  be  powerless  to 
balance  the  effect  on  total  output,  while  such  decline  may  produce  rising 
per  capita  figures  without  any  "progress"  in  our  sense,  possibly  even  with 
a  negative  one.  Finally,  we  should  take  account  of  variations  in  the 
amount  of  voluntary  leisure,  which  may  be  and  undoubtedly  often  is 
one  form  of  taking  increased  real  income  in  the  Fetter-Fisher  sense: 

1  For  the  United  States,  we  find,  assuming  the  population  to  reach  the  figure  of  about 
160  millions  in  1978  and  production  to  increase  at  3  per  cent  compound  interest,  that  income 
per  head  would  in  that  year  be  about  $2,800,  purchasing  power  of  1928,  when  income  per 
head  was  about  $700. 

2  The  idea  is  an  old  one  and  has  taken  several  different  forms.     But  we  will  mention 
only  Professor  S.  Kuznets,  Secular  Movements  in  Production  and  Prices,  and  Mr.  A.  F. 
Burns,  Production  Trends  in  the  United  States  since  1870. 


498  BUSINESS  CYCLES 

in  this  sense  the  reduction  of  working  hours  is  one  of  the  most  significant 
"products"  of  economic  evolution.  All  this  illustrates  the  difficulties  of 
speaking  of  retardation  if  the  object  is  to  measure  capitalist  performance 
with  a  view  to  deciding  whether  or  not  there  are  "hitches  "  inherent  to  the 
working  of  the  profit  economy  that  tend  to  thwart  mankind's  economic 
possibilities.  Further  illustration  is  afforded  by  the  following  examples 
of  actual  or  possible  causes  of  retardation.1 

First,  we  will  restate  statistical  reasons,  all  of  them  previously  men- 
tioned. As  in  the  case  of  indices  of  price  level,  we  have  to  accept  the 
fact  that  any  index  of  total  or  of  consumers'  goods'  output  will  in  general 
be  influenced  by  individual  and  group  developments.  In  the  case  of  the 
production  index,  this  spells  downward  bias:  new  commodities  and  fin- 
ished commodities  are  inadequately  represented,  services  (and  voluntary 
leisure),  not  at  all,  while  commodities  that  are  in  the  process  of  being  dis- 
placed loom  large.  Improvements  in  quality,  particularly  in  durability, 
largely  escape,  and  they  may  in  many  cases  be  increasingly  important. 
The  quantities  of  raw  materials  and  semifinished  goods  which  form  the  bulk 
of  what  enters  into  indices  of  output  are,  in  the  course  of  technological 
progress  and  because  of  the  spreading  use  of  waste  or  scrap,  made  up  into 
increasing  quantities  of  finished  goods,  so  that  observed  retardation  may 
be  spurious  and  even  indicative  of  its  very  opposite.  Finally,  the  period 
1897  to  1913  covers  the  prosperity  phase  of  a  Kondratieff — a  fact  that 
would,  according  to  our  theory,  be  sufficient  to  produce  an  impression  of 
retardation  if  comparison  is  with  the  years  of  Kondratieff  depression  and 
recovery,  1870  to  1896. 

Second,  there  are  what  we  may  term  sociological  reasons.  Some  of 
them — which  may  account  for  some  slackening  in  entrepreneurial  effort 
and  cognate  phenomena — have  been  alluded  to  in  the  third  chapter  and 
will  cross  our  way  again  in  our  discussion  of  the  postwar  period.  But  it 
has  also  been  pointed  out  that,  in  part  at  least,  economic  evolution  grows 
increasingly  independent  of  the  typically  capitalistic  pattern  of  cultural 
values  and  motives.  In  any  case,  the  writer  believes  it  to  be  safe  to 
discard  this  group  of  factors  in  an  interpretation  of  American  and  German 
developments  during  prewar  times,  although  in  the  case  of  England  we 
have  seen  reasons  for  some  doubt  about  this. 

1  The  reader  should  compare  Chap.  IV  of  Mr.  Burns's  book  mentioned  in  the  preceding 
footnote  for  a  suggestive  discussion.  He  will  find  that  statistical  difficulties,  formidable  as 
they  are,  are  secondary  to  the  difficulties  of  interpretation  which  of  course  differ  widely 
according  to  the  purpose  in  hand  and  can  be  expected  to  yield  only  to  most  careful  analysis 
of  industrial  processes  in  detail.  Inferences  as  to  decreasing  returns  or  any  slackening  of 
rate  of  progress  in  efficiency  (as  measured  by  service)  may  easily  prove  misleading.  As  to 
the  immediate  statistical  result  of  Mr.  Burns's  investigation,  the  writer  agrees  with  the 
views  expressed  by  Professor  Mitchell  in  the  preface  to  that  book.  Compare  also  Pro- 
fessor Crum's  review  in  Quarterly  Journal  of  Economics  for  August  1934. 


PHYSICAL  QUANTITIES.     EMPLOYMENT  499 

Third,  we  have  the  economic  reasons.  What  eventually  may  turn 
out  to  be  the  most  fundamental  of  them  and  also  account  for  some  of  the 
sociological  ones — increasing  satisfaction  of  wants — can  hardly  have  been 
operative  in  our  period.  For  it  takes  more  for  it  to  assert  itself  than  the 
mere  fact  of  declining  marginal  utility  of  income,  and  of  the  many  subtle 
elements  that  enter  into  this  problem  none  can  be  identified  with  any 
certainty  unless  indeed  we  interpret  certain  cases  of  the  struggle  for 
shorter  hours  in  this  sense.  Nor  can  decreasing  returns,  in  the  sense  of 
increasing  difficulty  of  procuring  food  stuffs  and  raw  materials,  have 
played  a  major  role  for  the  period  as  a  whole.  Any  Kondratieff  pros- 
perity will  of  course  tend  to  display — up  to  a  considerable  level  of  general 
wealth — rising  prices  in  both  those  groups,  and  this  each  time  deceives 
many  economists  into  believing  in  a  secular  law  of  decreasing  returns. 
But  it  is  a  temporary  phenomenon  or,  at  all  events,  it  has  been  a  tem- 
porary phenomenon  during  our  period. 

Again,  it  is  precisely  the  increasing  abundance  of  foodstuffs  and  raw 
materials  within  the  period  that  has  by  some  economists  been  made  the 
basis  of  an  argument  that  would,  if  true,  tend  to  establish  retardation 
in  a  special  sense  and  at  the  same  time  explain  the  failure  to  come  true  of 
Marxian  predictions  about  the  mass  misery  which  was  to  result  from  the 
working  of  the  capitalist  mechanism.  It  is  held  that  there  would  have 
been  retardation  either  from  decreasing  returns  or  from  the  nature  of 
the  profit  economy  but  for  the  fact  that  the  opening  up  of  new  countries 
temporarily  put  decreasing  returns  out  of  operation,  and  that  it  is  bound 
to  set  in  as  soon  as  this  unique  opportunity  is  exhausted.  We  have 
noticed,  in  Chap.  I,  that  it  is  not  correct  to  consider  the  actual  opening 
up — as  distinguished  from  the  discovery — of  new  countries  as  a  factor 
external  to  the  economic  system.  The  processes  of  exploiting  a  new 
country  are  but  a  type  of  innovation  and  alter  the  data  of  the  economic 
process  in  no  other  way  than  do  other  innovations.  Hence,  their  effects 
are  properly  included  in  a  measurement  of  capitalist  performance.  Nor 
does  any  prognosis  of  retardation  follow,  for  it  is  only  one  of  many  types  of 
innovations — and  one  only  of  the  many  types  of  innovation  in  the  fields 
of  foodstuffs  and  raw  materials — the  possibilities  of  which  thus  become 
exhausted.  The  possibilities  of  each  individual  type  do  so,  of  course, 
but  nothing  can  be  concluded  therefrom  for  innovation  in  general. 

Most  of  us  seem  here  to  commit  a  mistake  in  handling  the  concept  of 
decreasing  returns.  In  its  proper  sense  it  applies  to  given  production 
functions  and  generally  stationary  conditions  plus  Growth  only.  In 
order  to  make  it  relevant  to  any  forecast  about  the  future  course  of 
production,  it  must  be  used,  as  indeed  it  has  been  used  already  by 
Ricardo,  in  a  different  sense,  namely  in  the  sense  that  action  of  the  "static 
law"  of  decreasing  return  will  indeed  be  interrupted  by  innovation  but 


500  BUSINESS  CYCLES 

that  the  latter  will  be  powerless  to  compensate  its  effects  in  the  long  run 
— that,  as  it  were,  there  is  a  law  of  decreasing  returns  from  successive 
innovations.  And  in  this  sense  the  statement  is  entirely  unwarranted. 
Owing  to  their  unpredictability,  future  innovations  may  be  less  or  more 
conducive  to  increase  of  means  of  satisfying  wants  than  were  previous  ones. 
"The  great  things  may  be  done"  but  they  may  equally  well  be  still  to 
come :  while  the  earth  can  be  surveyed  and  better  opportunities  (relative 
to  each  given  state  of  technique)  for  its  exploitation  jnay  be  taken  up 
first,  so  that  inferior  ones  only  are  left  for  the  future,  the  world  of  possible 
innovation  cannot  be  mapped  out.1  Postwar  agricultural  history  testifies 
conclusively  on  this  point. 

C.  The  Cyclical  Behavior  of  the  Physical  Volume  of  Production. — 
It  will  be  recalled,  first,  that  our  theoretical  expectation  as  to  the  behavior 
of  total  output  was  the  net  result  of  our  theoretical  expectations  as  to 
the  behavior  of  two  constituent  groups  centering  respectively  in  pro- 
ducers' goods  and  the  commodities  that  enter  into  the  budgets  of  house- 
holds and,  second,  that  it  changed  with  successive  approximations. 
What  that  net  result  will  be  in  the  case  of  any  given  index  cannot  be  pre- 
dicted unless  we  know  what  commodities  are  included  and  how  it  is' 
constructed.  This  applies  with  greater  force  here  than  in  the  case  of  the 
price  level,  because  total  output  is  not  a  definite  real  thing.  The  com- 
parative steadiness  (in  the  sense,  say,  of  "good"  fit  to  the  material  of  an 
expression  of  the  form  y  —  ax1)  of  our  output  series  is  not  in  strict  theory 
primarily  due  to  steadiness  of  its  individual  constituents,  nor  simply 
to  the  smoothing  effect  on  random  fluctuations  in  constituents  of  combina- 
tion into  an  aggregate  (though  this  effect  will  also  be  present,  of  course), 
but  to  different  rhythms  in  the  two  groups  of  constituents.  These 
rhythms,  already  somewhat  toned  down  in  comprehensive  composites 
of  producers'  and  consumers'  goods,2  are  further  reduced  in  amplitude 
by  a  number  of  factors  mentioned  in  Chap.  IV,  which  we  will  presently 
restate. 

Our  expectation  that  total  output  will  increase  through  all  phases  of 
the  cycle,  "deep"  depression  alone  excepted,  is  derived  as  follows — the 
exception  hardly  ever  extending  over  the  whole  of  the  depressive  phase, 

1  Professor  Kuznets,  op.  cit.t  in  fact,  assumes  the  existence  of  such  a  law  of  decreasing 
returns  of  economic  progress,  and  has  tried  to  establish  it  by  analyzing  the  effects  of 
successive  innovations  in  technology.     The  partial  success  of  this  analysis  is,  however, 
merely  due  to  the  fact,  noticed  above,  that  in  every  industry  innovation  tends  to  exhaust 
itself  for  the  time  being  and  to  taper  off  into  comparatively  unimportant  improvements  of 
the  induced  or  completing  type. 

2  For  obvious  reasons,  durable  consumers'  goods  will  display  a  tendency  to  conform  to 
the  behavior  of  equipment  goods.     We  shall  repeatedly  refer  to  this  fact,  which  tends  to 
accentuate  fluctuations.     So  does  the  practice,  widely  observable,  of  putting  off  the  replace- 
ment of  equipment  in  depression  or  even  concentrating  it  in  prosperity. 


PHYSICAL  QUANTITIES.    EMPLOYMENT  501 

since  it  is  due  to  panics  and  vicious  spirals,  which  as  a  rule  do  not  last 
more  than  one  year.1  As  for  equipment  industries  and  their  satellites, 
expectation  from  our  model  does  not  differ  from  accepted  opinion  and 
only  provides  explanation  of  a  fact  universally  admitted  and  sometimes 
even  overstressed.2  Quantity  taken  should  increase  in  positive  phases 
and  decrease,  or  increase  less,  in  negative  phases.  Quantity  produced 
may  be,  and  undoubtedly  often  is,  particularly  in  the  course  of  the  shorter 
cycles,  interfered  with  by  exports  and  imports  and  other  circumstances, 
but  should  still  move  substantially  the  same  way.  Of  course  it  is  neces- 
sary to  take  account  first  of  the  interference  of  cycles  with  each  other; 
second,  of  the  fact  that  if  entrepreneurs9  demand  for  equipment  slackens 
— or,  in  strict  theory,  is  absent  in  three  of  the  four  phases — the  demand 
for  equipment  that  is  induced  by  entrepreneurial  activity  not  only  does 
not  cease,  but  actually  concentrates  in  recession  and  revival;  and,  third, 
of  Growth,  particularly  in  the  Kondratieff.  Inasmuch  as  we  may  take 
pig  iron  as  a  representative  of  the  equipment  group,  our  chart  of  pig-iron 
consumption  may  reasonably  be  said  to  verify  that  expectation.  Also, 
the  chart  of  rates  of  change  should  now  be  compared  with  the  chart  of 
rates  of  change  of  total  output. 

We  shall  expect  wholesale  prices3  to  precede  the  output  of  pig  iron 
or,  for  that  matter,  of  equipment  goods — at  least,  in  the  shorter  cycles — 
not  only  because  transactions  precede  production  and  because  of  the 
effect  of  speculative  anticipation,  but  also  because  wholesale  prices  will  in 

1  As  in  the  case  of  the  price  level,  we  shall  associate  them  with  the  two  short  cycles, 
because  in  the  long  and  gentle  sweep  of  the  Kondratieff  they  would  hardly  arise.     This 
does  not  mean  that  the  Kondratieff  has  nothing  to  do  with  them:  conditions  of  the  under- 
lying cycles  always  influence  events  in  the  course  of  the  superimposed  ones. 

2  We  may  illustrate  that  received  opinion,  first  effectively  preached  by  Tugan-Baranow- 
sky,  by  a  quotation  from  Marcel  Lenoir,  Etudes  sur  la  Formation  et  le  Mouvement  des 
Prix,  1913  p.  77,  of  the  implications  of  which  we,  of  course,  do  not  approve:  "la  cause  pro- 
fonde  des  fluctuations  periodiques  de  la  vie  economique  .  .  .  est  la  demande  intermittante 
de  capitaux  fixes  necessaires,  de  temps  en  temps,  pour  la  refection  et  1' augmentation  de 
1'outillage  economique."     If  refection  and  augmentation  were  all,  it  would  not  be  easy  to  see 
why  demand  for  "fixed  capital"  should  be  intermittent.     But  there  is  almost  universal 
agreement  about  the  fact. 

3  We  shall  not  be  so  sure  about  the  price  of  pig  iron  itself.     In  strict  theory,  this  price 
should  not  increase  in  revival,  since  the  demand  for  new  equipment  for  new  purposes  does 
not  set  in  until  the  end  of  that  phase.     But  both  replacement  demand  and  demand  from 
investment  opportunity  created  by  innovation  in  the  previous  prosperity  will  assert  them- 
selves in  recovery.     On  the  other  hand,  iron  and  steel,  being  cyclical  industries  and  holding 
a  reserve  of  capacity,  need  not  react  by  increasing  their  prices  at  first,  i.e.,  in  revival,  or 
even  at  the  beginning  of  prosperity.     The  reader  will  find  all  this  very  complicated.     If 
thereby  he  means  that  it  would  be  more  pleasant  if  reality  were  simpler,  the  writer  can  only 
heartily  agree  with  him.     If  he  thereby  means  to  record  an  objection  to  our  analytic 
schema,  the  writer  does  not  agree.     He  is  unable  to  see  any  merit  in  a  theory  yielding  simple 
and  clear-cut  propositions  which  do  not  fit  facts. 


502  BUSINESS  CYCLES 

general  have  to  make  up  for  a  fall  in  depression  and  hence  display  an 
increase  in  revival.  According  to  the  method  of  "lags  of  maximum  cor- 
relation" (Review  of  Economic  Statistics  1919,  p.  184,  et  seq.)  this,  in  fact, 
is  so.  If  indices  of  producers'  goods  consist  wholly  or  mainly  of  equip- 
ment goods,  the  same  will  hold  true  for  them.  Of  course,  no  causal  role 
of  prices  or  of  the  mechanism  of  money  and  credit  follows  from  that. 

But  our  expectation  as  to  finished  consumers'  goods  and  their  satel- 
lites, which  finally  boiled  down  to  expecting  that  their  output  should 
increase  more  in  recession  and  revival  than  in  prosperity,  runs  counter 
not  only  to  public  opinion,  but  also  to  the  opinion  of  most  students  of 
cycles.  In  its  "pure"  form,  our  model  yields  the  expectation  that  in 
prosperities  the  output  of  producers'  goods  should  at  first  increase  at  the 
expense  of  the  output  of  consumers*  goods.  The  latter  should,  even 
absolutely,  decline,  and  if  the  innovations  in  question  be  of  the  type  of 
railroad  construction,  no  fully  compensating  increase  of  the  former 
might  show  for  years.  In  general,  however,  new  products  will  be  released 
as  prosperity  wears  on,  their  impact  being  part  of  the  mechanism  that 
eventually  turns  prosperity  into  recession.  Hence  decrease  could  show 
only  in  a  segment  between  recovery  and  the  later  stages  of  prosperity. 
But  the  avalanche  of  consumers'  goods  and  their  satellites  should  come 
in  recession.  Their  output  should  continue  to  increase,  barring  panics 
and  spirals,  in  depression  and  display  the  strongest  increase  in  recovery. 

But  this  principle  holds  strictly  true  only  if  prosperity  starts  from 
perfect  equilibrium  in  perfect  competition.  This  not  being  so  in  reality, 
business  will  respond  to  brisk  demand  by  almost  immediate  expansion  of 
total  production  and  no  decrease  in  consumers5  goods  production  need 
occur  in  prosperity.  The  presence  of  cyclical  industries  acts  especially 
strongly  in  the  same  direction.  Moreover,  the  facts  that  we  refer  to 
under  the  heading  of  Growth  must  blur  the  picture.  The  population  of 
the  United  States  increased  from  1839  to  1915  at  a  compound  interest 
rate  of  2.28  per  cent  per  annum,1  and  we  may  assume  that  for  most  of 
that  period  no  tendency  toward  diminishing  returns  exerted  appreciable 
effects.  This  alone  would  be  sufficient  to  produce  an  increase  of  physical 
production  during  prosperity  phases  and  to  wipe  out  the  effects  of  any 
tendency  in  the  opposite  direction  such  as  could  be  expected  from  the 
working  of  our  mechanism.  But  while  this  tendency  will  therefore  gener- 
ally fail  to  show,  dips  due  to  the  breakdown  of  the  Secondary  Wave  and 
to  the  other  depressive  factors  will  almost  always  show.  They  influence 
both  people's  opinions  about  the  nature  of  downgrades  and  the  statistical 

1  It  should  also  be  taken  into  account  that  all  the  time  the  average  age  of  the  population 
was  increasing.  It  was  around  seventeen  years,  1790  to  1810  (white  males  only).  It  was 
eighteen  by  1820  (whites  only),  and  twenty-three  years  by  1910  (all  persons).  See  Census 
of  1920,  vol.  II,  p.  148. 


PHYSICAL  QUANTITIES.    EMPLOYMENT  503 

picture.  It  should  be  observed  that  they  are  likely  to  distort  a  short 
cycle  more  than  a  long  one,  in  the  course  of  which  there  is  time  for  things 
to  straighten  themselves  out  into  their  true  form.  We  shall,  hence, 
expect  the  facts  to  conform  best  to  the  above  expectation  in  the  case  of 
the  Kondratieff,  less  so  in  the  case  of  the  Juglar,  least  of  all  in  the  case 
of  the  Kitchin  cycle.  A  spurious  deviation  from  expectation  should  also 
be  kept  in  mind,  in  this  as  in  all  other  cases.  If  there  are  four  phases 
and  depression  contains  a  span  of  actual  fall  of  output,  then  revival  will 
be  the  phase  in  which  what  would  happen  in  the  downgrade  of  a  two-phase 
cycle  will  be  most  clearly  in  evidence.  But  most  writers  date  cycles 
from  troughs,  and  to  them  it  must  of  course  seem  as  though  we  were 
contradicting  an  obvious  fact. 

As  stated  above,  we  must  in  this  case,  as  in  all,  read  any  statement 
with  reference  to  the  presence  of  three  cyclical  movements  that  super- 
impose themselves  on,  and  interfere  with,  one  another.  Although  this  is 
more  difficult  to  take  account  of,  because  the  span  covered  by  material 
at  all  reliable  is  so  short,  it  thereby  loses  nothing  of  its  importance. 
Investigation  into  the  details  of  each  small  wave  would  be  necessary  in 
order  to  illustrate  it.  On  the  downgrade  of  a  Kondratieff,  the  increase 
in  consumers'  goods  incident  to  its  mechanism  will  tend  to  overshadow 
any  opposing  tendency  of  opposite  phases  of  the  shorter  cycles.  But 
this  is  not  all.  It  must  be  taken  into  account  also  that,  as  in  such  a 
downgrade  everything  is  technically  and  commercially  prepared  for 
expansion,  this  expansion  is  particularly  likely  to  show  under  the  influence 
of  increasing  expenditure  induced  by  prosperities  of  the  shorter  cycles, 
while  their  depressions,  notably  if  coinciding,  are  likely  to  spell  panic  and 
paralysis  of  business  and  so  to  inhibit  the  release  of  the  products  of  the 
new  industrial  apparatus,  amid  general  complaints  about  price  wars, 
cutthroat  competition,  and  overproduction.  On  the  upgrade — witness 
the  "hungry  forties"  and  conditions  around  the  turn  of  the  century — 
the  sweep  of  the  Kondratieff  intensifies  the  tendencies  of  prosperity 
phases  of  the  shorter  cycles  and  weakens  those  of  their  recessions  and 
revivals.  An  example  showing  how  important  it  is  to  attend  to  that 
phenomenon  is  afforded  by  the  indubitable  fact,  noticed  in  the  preceding 
section,  that  from  material  covering  1870  to  1913  it  is  possible,  for  all 
three  countries  and  many  others,  to  show  the  presence  of  a  long-time 
tendency  toward  declining  rates  of  increase  in  the  production  of  consumers' 
goods.  As  mentioned  before,  this  is  simply  due  to  the  circumstance  that 
the  period  1870-1897  happens  to  lie  in  what  first  was  a  depression  and 
then  a  recovery  phase  of  a  Kondratieff,  while  the  years  1897  to  1913 
cover  the  prosperity  and  only  a  few  years  of  the  recession  phase  of  a  new 
Kondratieff. 


504  BUSINESS  CYCLES 

Returning  to  charts  XV  and  XVI,  it  seems  reasonable  to  say  that  the 
theory  submitted  is  borne  out  by  the  behavior  of  total  industrial  produc- 
tion, l  inasmuch  as  the  theory  affords  a  rational  explanation  of  the  steadi- 
ness of  that  behavior.  For  if  this  theory  were  not  correct,  it  would  be 
extremely  unlikely  that  longer  periods  which  we  identify  as  nonprosperities 
should  display  on  the  whole  at  least  as  high  a  rate  of  increase  as  periods 
of  prosperity,  and  sometimes  a  higher  one.  And  if  alternative  theories 
were  correct  which  associate  increase  of  output  primarilynvith  prosperities 
and  decrease  or  smaller  increase  with  negative  phases,  it  would  be  equally 
unlikely  that  this  should  not  show  at  all  in  the  Kondratieff,  which  so 
strongly  displays  the  cyclical  variations  in  other  series.  Even  indications 
pointing  to  the  effects  to  be  expected  from  our  pure  model  are  not  entirely 
wanting.  If  the  reader  will  draw  straight  lines  through  the  intervals 
1898  to  1913  and  1858  to  1897  on  Chart  XIV,  he  will  find,  as  has  been 
mentioned  already,  that  for  all  three  countries  the  gradient  of  the  first 
line  is  smaller  than  that  of  the  second.  And  Professor  Mills' statement, 
(Economic  Tendencies,  p.  244)  to  be  noticed  again  in  our  discussion  of  the 
postwar  period,  undoubtedly  lends  support  to  the  opinion  that  in  the 
third  Kondratieff  output  actually  increased  in  recession  at  a  higher  rate 
than  in  prosperity.  Something  like  this  seems  also  to  be  true  of  the 
English  index  from  1820  (if,  on  account  of  the  Napoleonic  wars  and  their 
aftermath  it  is  permissible  to  begin  with  that  year  instead  of  some  date 
around  1800)  to  1842,  as  compared  with  1842  to  1858. 

Comparison  of  the  behavior  of  equipment  and  of  consumers'  goods, 
presented  on  Chart  XVII,  brings  out,  first,  some  familiar  features,  notably 
greater  amplitude  of  fluctuations  in  the  former.  But  a  lead  of  equipment 
goods  is  not  particularly  in  evidence.  There  is  no  reason  why  it  should 
be,  if  we  look  at  peaks  and  troughs — consumers'  goods  may  even  be 
expected  to  recover  more  quickly  from  troughs,  and  this  shows  in  several 
instances.  But  even  apart  from  this,  lead  and  lag  in  the  short  run  are 
largely,  from  the  standpoint  of  our  process,  a  matter  of  chance  or  rather 
of  peculiarities  of  individual  situations  that  have  little  fundamental 
significance.  Disregarding  questions  of  lead  and  lag,  however,  the  cyclical 
variation  in  the  relation  between  equipment  and  consumers'  goods 
shows  clearly  enough.  It  does  so  especially  in  the  three  lines  at  the 
bottom  of  the  chart,  which  are  particularly  useful  for  the  purpose  of 

1  There  is  really  no  such  thing  as  a  theory  of  total  output  as  such,  in  the  full  sense  of  the 
word,  for  this  would  imply  that  we  are  able  to  represent  it  as  an  explicit  and  uniquely  deter- 
mined variable  in  function  of  some  independent  variables  of  the  same,  i.e.,  the  systematic, 
class — such  as  price  level,  interest  rate,  quantity  of  circulating  medium.  Our  whole  analy- 
sis shows  this  to  be  impossible  and  any  theory  that  attempts  it,  to  be  a  sham.  What  we 
mean  above  is  simply  theoretical  expectation  as  to  behavior  of  output  withiu  a  process  that 
simultaneously  shapes  all  the  variables, 


PHYSICAL  QUANTITIES.     EMPLOYMENT 


505 


studying  Juglars.  Only  the  English  series,  characteristically  different 
from  the  other  two,  behaves,  if  anything,  contrary  to  expectation  as*  to 
the  Kondratieff,  and  shows  hardly  anything  except  a  short  irregular  wave 


PRODUCERS'  GOODS 


CONSUMERS'  GOODS 


PRODUCERS' GOODS 


CONSUMERS'  GOODS 


RATIO 


UNITED  STATES 


GREAT  BRITAIN 


UNITED  STATES 


JREAT  BRITAIN 


GERMANY 


GREAT  BRITAIN 


I8GO  1865  1870  1875  1830  1885  1890  1895  1900  1905  1910     1913 

CHART  XVII. — Industrial  equipment  and  consumers'   goods   (see  Appendix,  p.  1056). 

and  a  trend  easily  explainable  on  Boehm-Bawerk's  theory.  It  gives 
some  support  to  the  view  that  during  the  period  covered  the  stronger 
cyclical  impulses  came  to  the  English  organism  from  other  countries. 


506 


BUSINESS  CYCLES 


The  German  ratio  is  very  interesting.  It  rises  at  first  through  what  is 
distinctly  abnormal  behavior  of  consumers'  goods  at  the  beginning  of  the 
sixties,  then  goes  on  almost  level  (still  marking  Juglars,  however)  to 
1875,  jerks  up  after  a  dip  in  1877  (which  though  a  year  later  than  in  the 
United  States,  is  yet  an  early  indication  of  the  new  Juglar),  to  continue 
on  the  same  level  or  slightly  falling  to  1894,  after  which  the  new  Kon- 


A         > 

Jk  \  rf 
/i  \  \  /  / 
••    " 


n 


/BASIC  PRODUCTION 


1870          7S          1880          85  1890          95  1900        05  1910      13 

CHART   XVIII. — United  States  consumption  and  production  (see  Appendix,  p.  1057). 

dratieff  is  exceedingly  well  marked.  The  two  levels  might  be  associated 
with  depression  and  recovery  phases  of  the  second  Kondratieff.  The 
American  ratio  is  the  most  lively  one.  It  marks  the  Juglars  in  the  ratio 
with  a  clearness  that  leaves  little  to  desire  and  even — a  point  that  the 
reader  should  verify  carefully  because  there  is  plenty  of  room  for  differ- 
ence of  opinion  on  this — the  Kitchins.1  The  rush  to  a  new  level,  which 
is  not  lost  again,  stands  out  well  at  the  beginning  of  the  third  Kondratieff. 
Chart  XVIII  illustrates  this  still  better . 

There  is  some  point  in  comparing,  before  we  leave  the  subject, 
variations  in  output  with  variations  in  the  price  level.     We  recall  at  the 

1  The  Kitchins  can  be  best  observed,  for  all  countries,  on  the  chart  of  rates  of  changes. 


PHYSICAL  QUANTITIES.     EMPLOYMENT  507 

outset  that  positive  association  of  the  latter  with  the  cyclical  phases,  or 
at  least  three  of  them — rise  in  prosperity,  fall  in  recession,  and  depression 
— is  normal  in  the  sense  that  the  working  of  the  model  tends  to  produce  it, 
but  that  we  had  to  qualify  that  expectation  in  various  ways — notably 
with  respect  to  the  interference  of  cycles  with  each  other — even  within 
the  pure  theory  of  the  process.  We  have  now  to  qualify  further  by  taking 
account  of  what  has  been  said  about  the  behavior  of  output.  Never- 
theless, the  main  contour  is  as  required  by  our  schema.  The  reader 
should  look  at  the  three  pulse  charts  and  allow  himself  to  be  impressed 
with  the  two  great  result  trends  of  capitalist  evolution,  the  downward 
trend  of  prices  and  the  upward  trend  of  quantities,  both  products  of  the 
cyclical  process.  Although  we  must  not  simply  attribute  the  first  to  the 
second — they  are  both  elements  in  a  process  that  contains  many  other 
variables — we  may  take  the  picture  to  represent  one  of  the  outstanding 
features  of  that  process.  At  least,  if  business  activity  be  measured  by 
industrial  production,  it  would  be  patently  absurd  to  associate  it  either 
exclusively  or  primarily  with  periods  of  rising  price  level.  We  might 
even  use  the  fact  of  the  fall  (both  the  fall  in  each  Kondratieff  downgrade 
and  the  resulting  secular  fall)  of  the  level,  together  with  the  fact  that 
media  of  circulation  continued  to  increase,  for  a  short-cut  method  of 
proving — although,  to  be  sure,  this  would  not  be  quite  correct — the 
proposition  that  the  downgrade  plus  recovery  is  the  time  of  harvesting 
the  fruits  of  innovation  in  the  form  of  increased  output  and  that  given 
the  behavior  of  the  other  factors  as  it  actually  is — but  this  is  a  method  of 
reasoning  that  the  writer  always  highly  disapproves  of,  if  he  finds  some- 
body else  using  it — it  is  the  release  of  those  fruits  that  turns  the  price 
level.  Inspecting  for  instance  the  British  chart,  we  find  level  and 
output  rising  together,  with  what  we  might  imagine  would,  without  the 
French  War,  have  been  fairly  similar  gradients  up  to  1800  (when  output 
turns  within  a  short  cycle  before  prices),  where  we  suspect  the  turn  of 
prices  would  have  come  without  the  Napoleonic  wars.  Prices  rise  with 
interruptions  to  the  peak,  which  in  our  series  comes  in  1814;  output 
winds  up  steadily  through  its  Kitchins.  Then  prices  come  down  to  1820 
(the  Kitchin  in  output  that  corresponds  to  the  hump  in  prices  in  1817— 
1818  looking  not  very  different  from  its  predecessors),  and  output  con- 
tinues to  increase  in  utter  disregard  of  all  theoretical  argument  that  has 
been  put  forth  to  the  contrary  by  so  many  economists.  Then  prices 
sweep  down — output  sweeps  up  at  an  increasing  rate  to  1825  and  turns 
into  the  crisis  of  that  year  before  prices.  And  after  that,  it  continues  to 
increase — if  anything,  less  during  the  period  of  rising  prices  (1849  to 
1873)  than  it  did  in  the  preceding  period  of  falling  prices.  It  rises  some- 
what less  than  we  should  expect  from  1873  to  1895,  but  still  at  about  the 
same  rate,  in  spite  of  the  prolonged  fall  in  prices.  And  it  rises  somewhat 


508 


BUSINESS  CYCLES 


more  than  we  should  expect,  with  rising  prices  from  1895  to  1913,  but 
again  at  about  the  same  rate. 

In  the  case  of  Juglars,  it  happens  much  oftener  that  falling  prices  are 
associated  with  dips  in  output — precisely  in  depressions  or  "  crises  "  such 
as,  in  England,  in  1793  and  1794,  then  in  1884,  1885, 1886,  1892,  and  1893, 
in  1903  and  1904,  and  on  other  similar  occasions,  though  by  no  means 
consistently  so.  Prosperity  phases  repeatedly  rise  from  falling  levels, 


Per  Cent 

+40  r 


+30 
+20 
+10 


WHOLESALE  PRICES 


V\l 


UU4i 

r/%  «T /  v    * 


i 


-10 
-20 
-30 
+30 
+20 
+  10 


U 


\l 
v 


A 


BANK  CREDIT 
ACCOUNTS 
CIRCULATION 


-10  - 
+200- 
+100- 


BERLIN  MARKET  DISCOUNT  RATE 


-100- 
+20- 


INDUSTRIAL  PRODUCTION 


~IO!860         1865         1870         1875         1880         1885         1890          1895         1900         1905          1910    1913 
CHART  XIX. — German  prewar  pulse,  rates  of  percentage  change  (see  Appendix,  p.  1057). 

but  only  on  Kondratieff  downgrades,  as,  for  example,  the  English  pros- 
perity of  the  twenties  of  the  nineteenth  century,  preceding  the  crisis  of 
1825,  or  the  English  and  German  prosperities  of  the  eighties,  which, 
however,  proves  nothing  against  the  "normality"  of  increase  of  price 
level  in  prosperities.  There  is  no  systematic  association  of  recession 
with  falling  output.  But  in  the  Kitchins  there  seems  to  be:  although 
instances  to  the  contrary  are  not  lacking,  the  evidence  for  positive  associ- 
ation of  output  with  the  Kitchin  phases  and,  on  the  whole,  also  with  price 
level,  supported  as  it  is  by  theoretical  considerations,  is  fairly  strong. 


PHYSICAL  QUANTITIES.     EMPLOYMENT  509 

We  present  on  Chart  XIX  the  rates  of  change  of  the  series  we  know 
already  from  the  German  pulse  chart.  Covariation  is  much  in  evidence. 
It  seems  to  be  approximately  instantaneous,  although  rate  of  change  of 
output  displays  a  tendency  to  precede.1 

D.  Employment  of  Labor. — This  is  the  only  case  of  employment  of 
productive  resources  with  which  we  shall  deal.  Investigation  of  the 
variations  in  the  degree  of  utilization  of  natural  resources  constitutes 
one  of  the  unfulfilled  desiderata  of  our  program.  As  regards  variations 
in  the  degree  of  utilization  of  plant  and  equipment,  we  must,  for  lack  of 
the  kind  of  information  we  need  for  our  purpose,  confine  ourselves  to 
repeating  that,  the  more  successfully  progressive  a  community  is,  the 
more  underutilization  or  excess  capacity  of  plant  and  equipment  it  must, 
other  things  being  equal,  display,  both  because  rapid  progress  means 
violent  disturbance  and  correspondingly  violent  upsets,  and  because  the 
more  rapid  the  advance,  the  greater  the  percentage  of  scmiobsolete  and 
obsolescent  plant  and  machinery  will  be  which  is  not  yet  definitely  dis- 
carded by  firms  or  dropped  from  statistical  lists.  Material  concerning 
variations  in  the  employment  of  labor  is  very  scarce  and  defective  for 
the  whole  of  the  prewar  period.  For  Germany  we  have  indications  only 
— mainly  from  the  nineties  on — which  may  easily  mislead.  The  decen- 
nial United  States  census  data  are  no  good  for  our  purpose.  The 

1  Professor  Irving  Fisher  (Our  Unstable  Dollar  and  the  So-Called  Business  Cycle, 
Journal  of  the  American  Statistical  Association  for  June,  1925)  has  found  by  the  method  of 
lag  distribution  that  maximum  correlation  between  W.  M.  Persons'  index  of  the  volume 
of  trade  and  variation  in  the  rate  of  change  of  the  price  level  is  obtained  if  we  lag  the  former 
by  seven  months.  If  any  trust  can  be  placed  in  this  lag,  which  is,  as  it  were,  the  center  of 
gravity  of  the  "effects"  of  the  change,  and  shows  the  maximum  correlation  of  those  indices 
for  the  period  from  August  1915  to  March  1928,  it  would  be  rather  tempting  to  urge  that 
a  lag  of  that  length  would,  in  fact,  roughly  correspond  to  the  distance,  in  a  perfectly  regular 
Kitchin  cycle  lying  on  a  Kondratieff  downgrade,  between  the  peak  of  price  level  (at  or 
near  the  end  of  the  prosperity  phase)  and  the  peak  of  output  (at  the  end  of  recession).  We 
will  not,  however,  insist  on  this.  The  idea  of  trying  to  establish  causal  connection  in  this 
way  seems  to  us  a  mistaken  one.  No  functions  connecting  two  variables  only  will  ever  do 
in  this  field.  And  what  covariation  there  is  seems  to  us  much  better  described  and  inter- 
preted as  an  imperfect  instantaneous  one. 

Professor  Pigou,  Industrial  Fluctuations,  1st  ed.,  p.  28,  makes  the  unguarded  state- 
ment that  industrial  expansions  are  "invariably"  characterized  by  rising  prices  and  indus- 
trial depressions  by  falling  prices.  In  order  to  keep  this  statement  from  being  less  true  than 
the  opposite  one  would  be,  it  must  be  confined  to  what  Professor  Pigou,  in  fact,  calls  "nor- 
mal" cycles,  i.e.y  roughly  speaking,  Kitchins.  It  could  include  Juglars  only  if  we  replace 
"expansions"  (i.e.,  of  output)  by  prosperities.  Even  so,  Professor  Persons'  criticism  in 
his  review,  Quarterly  Journal  of  Economics  for  1927-1938,  p.  672,  is  statistically  well 
founded.  But  neither  seems  to  perceive  the  true  relation  of  output  to  the  cyclical  phases 
in  prices,  or  to  realize  that  no  expectation  can  be  formed  either  a  priori  or  from  statistical 
regularity  as  to  lag  or  lead  between  output  and  prices.  We  quote  both  statement  and 
criticism  because  of  their  methodological  interest. 


510  BUSINESS  CYCLES 

Massachusetts  figures,  though  very  valuable  in  themselves,1  start  only 
in  1889.  We  shall  in  this  section  use  English  figures  only,  taking  unem- 
ployment percentage  as  an  index  of  the  variations  in  employment.  The 
figures,  which  start  in  1851,  present  a  sample  only — percentage  of  the 
number  of  members  of  trade  unions  reported  as  out  of  work2 — and  do  not 
take  account  of  part  time.  They  exclude,  however,  workmen  who  were 
idle  because  of  strikes  or  lockouts  or  illness  or  age.  The  series  has  been 
used  and  compared  with  others  (such  as  market  rate  of  discount  on  three- 
months  bank  bills,  pig-iron  consumption,  London  clearings,  Sauerbeck 
prices,  increase  in  bank  credits  outstanding,  rate  of  real  wages,  consump- 
tion of  meat,  beer,  etc.)  by  Professor  Pigou  in  Industrial  Fluctuations. 
In  order  to  save  space,  we  will  replace  evidence  that  we  ought  to  present 
by  a  reference  to  that  work,  as  well  as  to  Professor  W.  Persons'  comments 
previously  quoted. 

We  have  seen  that  there  is  no  unique  or  simple  relation  between 
employment  (number  of  hours  worked  per  week)  and  output,  and  that 
the  latter  is  not  proportional  to,  or  measured  by,  the  former.  This  is  a 
consequence  of  the  very  nature  of  economic  evolution  and  becomes 
obvious  as  soon  as  some  of  the  conditions  for  proportionality  are  stated : 
production  functions  would  have  to  be  invariant  in  time  and  relative 
prices  of  factors  would  have  to  be  constant.  Neither  can  possibly  be 
fulfilled  for  any  length  of  time,  such  as  the  period  of  a  Juglar.  But  both 
may  be  fulfilled  approximately  in  the  very  short  run,  for  which,  moreover, 
the  second  need  not  always  be  fulfilled,  because  adaptation  to  change  in 
relative  prices  of  factors  may  not  be  possible  within  it.  This  very  short 
run  may  extend,  although  it  cannot  be  relied  on  to  do  so,  to  the  span  of  a 
Kitchin,  which  will,  therefore,  in  this  respect,  present  a  picture  different 
from  that  of  the  longer  cycles.  It  should  also  be  observed  that  total 
employment  or  unemployment  would  in  any  case  be  unsatisfactory  as  an 
index  of  business  situations,  even  if  figures  were  as  exact  as  they  are 
rough,  and  even  if  it  were  possible  to  correct  them  for  all  the  circumstances 
that  change  their  significance  in  time  (attitude  of  workmen  to  unem- 
ployment and  to  relief,  geographical  and  industrial  mobility,  length  of 
working  day,  and  so  on).  For  variation  in  total  employment  or  unem- 
ployment is  the  net  result  of  what  actually  goes  on  in  industrial  regions 
and  trades — let  alone  concerns — and  tends  to  hide  differences,  which 

1  For  various  estimates  of  employment  and  unemployment  in  this  country,  including 
those  of  Berridge,  Hurlin,  Brissenden,  and  Douglas-Stinebower,  see  Professor  P.  H.  Doug- 
glas's  Real  Wages  in  the  United  States,  1890-1926,  1930,  to  which  the  present  writer  is 
much  indebted. 

2  Professor  Bowley  made  an  attempt  to  go  beyond  those  figures,  to  which  attention 
should  be  drawn.     See  The  Measurement  of  Employment :  an  Experiment,  Journal  of  the 
Royal  Statistical  Society  for  July  1912,  in  which  paper  he  also  shows  why,  as  a  measurement 
of  general  employment,  our  series  is  anything  but  trustworthy. 


PHYSICAL  QUANTITIES.    EMPLOYMENT  511 

from  our  standpoint  often  are  what  matters  most.  This  may  be  illus- 
trated by  the  case  of  seasonal  unemployment,  which  may  be  considerable 
and  yet  produce  but  modest  seasonal  amplitudes,  if  seasons  differ  with 
different  industries.1 

We  will  now  inspect  the  graph  of  our  series  (Chart  XX)  and  at  the 
outset  notice  the  negative  covariation  of  per  cent  unemployed  with  the 
rate  of  interest  (see  also  Pigou,  Industrial  Fluctuations,  Chart  facing 
p.  28),  which  stands  out  strongly  and  is  blurred  only  by  the  fact  that 
Kitchins  show  much  more  in  the  graph  of  the  latter.2  Cases  such  as 
the  one  in  1854,  which  was  obviously  due  to  the  dislocations  incident  to 
the  outbreak  of  the  Crimean  War,  do  not  amount  to  real  exceptions. 
For  the  purpose  of  analyzing  this  picture,  we  will  recall  or  introduce  the 
following  concepts  and  propositions. 

We  will  call  Normal  Unemployment  the  unemployment  that  would  at 
any  point  of  time  exist  if  the  system  had  already  reached  the  neighborhood 
of  equilibrium  toward  which  it  is  tending.  It  comprises  the  cases  of 
seasonal  unemployment  which,  however,  for  the  reason  just  mentioned, 
need  not  reveal  itself  completely  and  which,  as  pointed  out  in  the  first 
chapter,  may  display  a  special  trend  owing  to  technological  change, 
such  as  has  in  the  postwar  period  occurred  in  building,  or  to  change  in 
consumers'  habits.  It  also  comprises  cases  of  unemployment  due  to 
such  accidents  (for  example  accidental  destruction  of  factories)  as  will 
ordinarily  occur,  of  unemployability  (which  however  we  may  neglect,3 
since,  for  England  and  that  period,  trade  unions  were  not  likely  to  harbor 
many  such  cases)  and  of  unemployment  due  to  change  of  residence  or 
occupation  or  job.  Finally,  we  have  seen  that  imperfections  of  com- 
petition or  of  equilibrium  may  account  for  failure  of  the  system  to  absorb 
all  its  eligible  workmen.  This  is  the  only  meaning  we  can  attach  to  the 
phrase  frequently  used  in  Germany  but  in  a  different  sense,  Structural 
Unemployment.  With  reference  to  this  concept  of  normal,  we  define 
supernormal  and  subnormal  unemployment.  It  will  be  observed  that 
subnormal  unemployment  in  this  sense  becomes,  outside  of  the  precincts 
of  perfect  competition,  compatible  with  a  high  figure  of  unemployment, 

1  See  Charles  Saunders,  Importance  of  Seasonal  Variations  in  Employment  in  the  United 
Kingdom,  Economic  Journal  for  June  1935. 

2  For  that  reason,  correlation  between  the  two  series  would  not  be  very  good.     Even  if 
the  relation  be  judged  by  the  relative  number  of  opposite  year-to-year  directions  of  move- 
ments, the  result  is  not  particularly  satisfactory.     But  this  affords  only  another  instance 
of  the  failure  of  formal  methods  to  work  satisfactorily.     Everyone  will,  nevertheless,  be 
struck  by  the  reality  of  the  relation.     In  this  case  a  smoothing  device  would  suffice  to  bring 
it  out  formally. 

8  It  will  be  understood,  however,  that  unemployability,  extending  as  it  does  far  beyond 
technically  pathological  cases,  is,  for  other  purposes  than  ours,  not  only  not  negligible  but 
both  economically  and  socially  one  of  the  most  serious  of  all  the  problems  of  unemployment . 


512 


BUSINESS  CYCLES 


PHYSICAL  QUANTITIES.    EMPLOYMENT  513 

that  is  to  say,  with  a  state  of  things  in  which  many  people  fail  to  find  work 
that  they  are  both  qualified  and  willing  to  undertake  at  the  ruling  rate 
of  wages.  Supernormal  employment  includes  overtime,  subnormal 
employment  short  time;  but,  as  stated  above,  neither  can  be  taken 
account  of  for  our  period. 

A  fundamental  difficulty  must  be  noticed,  although  we  cannot  do 
much  about  it.  The  various  sources  that  contribute  to  any  given  sum 
total  of  unemployment  are  not  independent  and  their  effects  cannot  be 
separated.  In  particular,  the  cyclical  process  affects  them  all  and  cyclical 
variations  of  unemployment  are  affected  by  them  all.  This  is  true  also  of 
all  types  of  normal  unemployment  and  even  of  unemployability ;  for  the 
question  whether  a  given  individual  is  unemployable  or  not  will  be 
answered  differently  in  times  of  good  business  and  in  times  of  bad  busi- 
ness. Moreover,  the  same  accidental  disturbance  will  give  rise  to  differ- 
eiht  amounts  of  unemployment  or  to  unemployment  of  different  duration, 
according  to  the  cyclical  phase  with  which  it  happens  to  coincide  and  so  on. 
Hence,  our  concept  of  a  normal  percentage  of  unemployment  must  not 
be  taken  to  mean  an  independent  quantity  that  could  be  added  to  equally 
independent  quantities  of  other  kinds  of  unemployment,  but  simply  the 
percentage  of  workers  unemployed  which  would  exist  in  the  absence  of 
disturbances  of  equilibrium.  In  order  to  form  an  idea  about  its  numerical 
importance,  we  must  first  locate  neighborhoods  of  equilibrium :  an  average 
of  observed  values  would  obviously  be  quite  meaningless.  1897  is  a 
particularly  good  year  to  choose  since,  according  to  our  schema,  all  our 
cycles  then  passed  their  neighborhoods.  The  English  figure  for  that 
year  was  3.3  per  cent,  and  we  are  perhaps  not  very  far  off  the  mark  if  we 
consider  this  value  as  approximately  normal  for  England  at  that  time. 

If,  given  a  state  of  otherwise  perfect  equilibrium  in  a  perfectly  com- 
petitive system,  wage  rates  are  raised,  say,  by  public  authority,  above 
their  equilibrium  value,  a  definite  amount  of  unemployment  will,  other 
things  remaining  equal,  ensue,  which  we  will  call  Vicarious  Unemploy- 
ment. If  either  equilibrium  or  competition  be  imperfect,  we  may  still 
use  this  concept  to  designate  unemployment  due  to  deviation  of  wage 
rates  from  the  figure  at  which  normal  employment  would  be  attained, 
or — hence  our  term — unemployment  that  takes  the  place  of  adaptation  of 
wages  to  that  figure.  So  far  the  concept  is  perfectly  definite — though 
it  may  still  be  impossible  to  measure  it  numerically — and  simply  expresses 
a  certain  aspect  of  one  element  of  normal  unemployment,  provided  that 
state  of  things  lasts  throughout  the  period  under  consideration,  or  a 
particular  type  of  supernormal  unemployment,  if  it  is  but  temporary. 
But  we  shall  also  use  the  term  to  indicate  elements  of  unemployment 
that  could  be  removed  by  an  appropriate  change  in  wage  rates,  outside 
of  neighborhoods  of  equilibrium,  And  here  th§  goncept  is,  for  reasons 


514  BUSINESS  CYCLES 

that  will  become  apparent  in  our  discussion  of  the  behavior  of  wages, 
extremely  difficult  to  handle. 

We  have  above  included  within  normal  unemployment  such  cases  as 
will  currently  arise  from  accidents  to  firms  and  are  analogous  to,  say,  the 
current  rate  of  deaths  by  accident.  If  firms  suffer  economic  injury  to  a 
supernormal  degree — analogy:  deaths  from  an  epidemic — we  speak  of 
Disturbance  Unemployment.  Many  instances  of  the  unemployment 
we  observe  in  the  course  of  cycles  obviously  belong  here.  Revolutions 
or  liquidations  of  wars  afford  others.  From  the  standpoint  of  any  indi- 
vidual industry,  a  depressive  state  in  another  industry  also  gives  rise  to 
disturbance  unemployment.  A  particularly  important  class  of  cases 
consists  of  environments  in  the  process  of  decline  from  long-time  causes 
external  to  the  system.  A  country  being  cut  off  from  its  hinterland  by 
the  insertion  of  a  new  frontier  line  illustrates  what  is  meant.  In  this 
category  we  include,  when  dealing  with  any  single  country  by  itself,  the 
effects  of  innovations  that  develop  and  run  their  course  outside  of  it, 
for  instance,  the  effects  of  the  innovation  of  new  routes  of  trade  on 
Venice  after  the  sixteenth  century,  or  the  effects  on  modern  Europe  of  the 
rise  of  large-scale  industry  in  the  tropics. 

But  for  the  special  case  of  unemployment  arising  from  disturbance 
by  innovation  within  the  system  we  will  set  up  a  distinct  class,  to  be 
called  Technological  Unemployment.  This  term  taken  literally,  of 
course,  has  always  been  intended  to  cover  only  displacement  of  workmen 
by  machinery.  We  make  it  cover  a  much  wider  range  and  include  not 
only  the  effects  on  employment  of  every  kind  of  change  in  industry  and 
commerce — organizational  change,  for  instance — but  also  the  effects 
which  changes  have  on  employment  in  firms  or  industries  that  are  com- 
peted with  by  the  firms  of  industries  that  introduce  new  production 
functions.  Questionnaires  devised  to  find  out  from  workmen  reasons 
for  their  dismissal  can,  therefore,  never  bring  out  the  phenomenon  we 
mean  and  will  always  yield  results  that  understate  it.1  Yet  it  should  be 
clear  that  our  wide  definition  does  but  justice  to  the  phenomenon,  since 
all  it  does  is  to  extend  the  usual  meaning  of  the  concept  to  cases  funda- 
mentally identical  in  economic  meaning  and  causation  to  the  one  com- 
monly envisaged  under  that  heading.  It  is  obvious,  for  instance,  that  in 
the  case  of  replacement  of  a  carriage  by  a  motorcar,  the  coachman  will 

1  There  is  another  reason  for  this.  The  1930  Census  of  Unemployment  attempted  to 
ascertain  the  causes  of  dismissal  of  unemployed  workmen  and,  under  the  heading  of  Indus- 
trial Policy,  inserted  the  technological  cause  as  one  of  five  subheads.  According  to  the 
answers  given,  the  men  themselves  must  have  considered  all  these  five  subheads  of  negligi- 
ble importance.  A  little  reflection  will,  however,  show  that  only  in  a  minority  of  cases 
will  workmen  be  able  to  recognize  the  technological  change  responsible  for  their  dismissal. 
For  this  it  would  be  necessary  that  machines  be  introduced  in  an  existent  plant  under  the. 
eyes  of  the  workmen  and  that  dismissal  be  effected  immediately  after. 


PHYSICAL  QUANTITIES.    EMPLOYMENT  515 

be  technologically  unemployed  even  in  the  narrow  sense,  although  no 
machine  drives  his  horses  henceforth,  or  that  it  does  not  make  any  differ- 
ence whether  a  bookkeeping  clerk  becomes  unemployed  because  of  the 
introduction  of  a  calculating  machine  or  another  rationalizing  device,  or 
whether  a  cotton  picker  becomes  unemployed  because  of  the  introduc- 
tion of  a  cotton-picking  machine  or  because  cotton  is  being  eliminated 
by  the  competition  of  the  standard  fiber.  Since  every  kind  of  unemploy- 
ment will  induce  further  unemployment,  Secondary  Unemployment  in 
Mr.  R.  F.  Kahn's  sense  must  be  added  to  each. 

Now,  a  large  majority  of  economists  will  agree  in  believing  that 
vicarious  unemployment  contributes  something  to  the  total  by  which 
unemployment  varies  in  the  course  of  cycles  (Cyclical  Unemployment), 
although  they  do  very  much  disagree  about  the  importance  of  that 
contribution.  Our  own  views  about  this  will  be  developed  in  the  dis- 
cussion on  the  behavior  of  wages.  All  economists  will  agree  to  the 
statement  that  disturbance  unemployment  contributes  much.  Panics, 
spirals,  prosperities  which  facilitate  the  emergence,  and  recessions  which 
facilitate  the  breakdown,  of  fraudulent  or  ill-conceived  enterprise,  and 
all  of  the  phenomena  that  produce  the  rise  and  break  of  the  Secondary 
Wave  amply  account  for  that.  As  far  as  these  points  go,  all  we  have  to 
add  is  that  such  phenomena  would  follow  from  the  working  of  the 
cyclical  mechanism,  whatever  the  behavior  of  money  and  credit,  although 
autodeflation  accentuates  them  and  their  effect  on  employment.  But 
few,  if  any,  economists  realize  the  one  major  point  that  the  writer  wishes 
to  make.  They  have  a  habit  of  distinguishing  between,  and  contrasting, 
cyclical  and  technological  unemployment.  But  it  follows  from  our  model 
that,  basically,  cyclical  unemployment  is  technological  unemployment. 
For  vicarious  and  disturbance  unemployment  are,  in  the  main,  but 
understandable  incidents,  though  quantitatively  important  in  practice, 
which  we  could  abstract  from  without  thereby  blotting  out  any  essential 
contours.  Technological  unemployment,  however,  is  of  the  essence  of 
our  process  and,  linking  up  as  it  does  with  innovation,  is  cyclical  by 
nature.  We  have  seen,  in  fact,  in  our  historical  survey,  that  periods  of 
prolonged  supernormal  unemployment  coincide  with  the  periods  in  which 
the  results  of  innovations  are  spreading  over  the  system  and  in  which 
reaction  to  them  by  the  system  is  dominating  the  business  situation, 
as,  for  instance,  in  the  twenties  and  in  the  eighties  of  the  nineteenth 
century. 

It  further  follows  that,  like  profits,  technological  unemployment  is 
ephemeral.  It  might  nevertheless  be  ever  present,  but,  as  in  the  case 
of  profits,  every  individual  source  of  it  in  the  industrial  organism  tends 
to  exhaust  itself,  while  new  ones  emerge  periodically.  In  the  same  sense 
as  profits,  moreover,  it  may  be  called  frictional,  since  instantaneous 


516  BUSINESS  CYCLES 

adaptation  of  the  system  would  kill  it  at  birth.  The  reader  need  not 
fear  that  the  writer  harbors  any  intention  of  using  these  statements  in 
order  to  minimize  the  importance  of  the  phenomenon  or  the  sufferings  it 
inflicts.  But  it  should  be  noticed  that,  in  consequence,  the  primary 
long-run  interest  of  the  working  class  is  in  the  effects  of  innovation  on  the 
total  real  wage  bill  and  not  in  the  incident  variation  of  employment, 
which  is  but  an  element  of  the  mechanism  that  produces  the  changes 
of  the  former  and  can  be  separately  handled  by  public  policy. 

We  now  proceed  to  apply  this  analysis  to  the  task  of  explaining  the 
behavior  of  our  series.  Unemployment  is  one  of  the  primary  elements  of 
our  process  and,  though  entirely  consequential,  also,  like  everything  else, 
a  conductor  of  effects  which  exert  secondary  causal  influences  of  their  own. 
The  series  is  natural,  systematic,  and  obviously  also  cyclical.  But  our 
theory  does  not  suggest  any  reason  for  expecting  any  result  trend  in 
unemployment  percentage.  We  have  before  us  one  of  the  two  instances, 
among  the  major  elements  of  our  statistical  picture,  of  clean  cyclical 
series,  the  other  being  the  series  of  interest  rates.  The  reader  should 
carefully  reconstruct,  from  the  fourth  chapter,  every  step  of  the  reasoning 
that  leads  up  to  this  statement.  Absence  of  trend  in  percentage  of 
unemployed  workmen  would  mean,  in  the  case  of  a  population  constant  in 
number  and  age  distribution,  that  the  process  of  evolution  works  in  such 
a  way  as  to  absorb  all  the  cyclical  unemployment  it  creates,  technological 
and  other.  If  net  increase  in  number  of  population  (natural  increase 
minus  emigration  plus  immigration)  fails  to  produce  a  rising  trend,  this 
means  that  the  system  also  absorbs  the  current  increment  at  the  same 
terms,  i.e.,  subjecting  it  to  the  same  unemployment  percentage.  Absorp- 
tion of  emigrants  from  the  agricultural  sphere  belongs  to  the  first  and 
not  to  the  second  class,  even  if  those  "emigrants"  were  proprietors  before, 
or  for  other  reasons  not  technically  unemployed. 

A  glance  at  the  chart  seems  in  fact  to  verify  this  result.  We  observe 
no  obvious  trend,  and  it  is  clear  that  any  trend  that  could  possibly  be 
derived  by  formal  methods  would  be  too  weak  to  be  significant,  particu- 
larly if,  recalling  what  has  been  said  about  the  untrustworthiness  of 
individual  extremes,  we  do  not  attach  much  importance  to  the  peaks 
that  occur  in  1858  and  1879.  It  is  no  doubt  true  that  our  statistical 
data  cannot  be  relied  on  implicitly.  Among  other  things,  trade-union 
figures  do  not  reflect  juvenile  unemployment  and  the  variations  of  the 
delay  in  getting  first  jobs.  Unemployment  among  unskilled  and  unor- 
ganized workmen  may  have  behaved  in  a  different  way.  Some  absorption 
could  have  been  expected  to  occur  from  the  mere  effect  of  such  saving  as 
might  have  occurred  irrespective  of  our  process.  Moreover,  institutional 
changes,  insertion  into  the  system  of  additional  permanent  rigidities, 
emergence  at  any  time  of  imperfections  that  had  been  absent  before — for 


PHYSICAL  QUANTITIES.     EMPLOYMENT  517 

instance,  of  the  oligopolistic  type — could  have  produced  a  positive 
trend,  their  elimination  a  negative  trend,  so  that  the  basis  of  our  theory 
and  the  basis  of  our  statistical  finding  are  not  simply  congruent.  But  it 
is  not  likely  that  such  factors  should  have  combined  exactly  so  as  to 
produce  a  spurious  verification  of  our  thesis,  or  that  fuller  information 
would  substantially  affect  it  beyond  showing  up  a  number  of  particular 
deviations  due  to  particular  causes. 

As  we  shall  see,  cyclical  unemployment  has  within  the  epoch  under 
survey  been  absorbed  at  increasing  rates  of  real  wages.  It  is  this  fact 
that  lends  importance  to  both  our  result  and  our  observation,  which  would 
become  trivial  if  absorption  had  been  brought  about  by  downward 
revision  of  real  wages.  For  the  moment,  however,  we  will  merely  notice 
that,  exactly  as  unemployment  may  be  vicarious  to  a  fall  in  wages,  so  a 
rise  in  wages  may  be  vicarious  to  an  increase  in  employment.  As  in  the 
former  case  we  can,  theoretically  at  least,  express  an  increase  in  unem- 
ployment in  terms  of  the  equivalent  fall  in  wages  that  would  prevent  it, 
so  we  can  in  the  latter  case  express  an  increase  in  wages  in  terms  of  an 
equivalent  increase  in  employment  which  is  only  virtual,  of  course,  since 
there  was  no  corresponding  labor  force  in  existence.  Thus  reducing  the 
long-run  effects  of  the  cyclical  process  on  wages  to  effects  on  virtual 
employment,  we  arrive  at  a  negative  trend  in  percentage  of  virtual  unem- 
ployment, while,  if  absorption  had  been  brought  about  by  a  fall  in  wages, 
the  trend  of  virtual  unemployment  would  be  positive. 

Expectations  as  to  the  cyclical  behavior  of  the  unemployment  per- 
centage are  too  obvious  to  be  restated.  If  a  cycle  started  from  perfect 
equilibrium  in  perfect  competition,  increase  of  employment  in  prosperity 
could  show  only  by  overtime  if  at  all.  As  it  is,  it  partly  shows  by  over- 
time— hence,  imperfectly  in  our  series.  The  unemployment  of  deep 
depression  will  be  as  irregular  as  everything  is  in  that  phase.  It  will  also 
be,  owing  to  the  preponderance  of  what  we  have  called  disturbance,  of  a 
sufficiently  distinct  nature  to  warrant  a  distinct  name :  Depression  Unem- 
ployment. Revival  should  bring  employment  up  to  normal  and  actually 
does  so  more  frequently  than  observers  assume  who  neglect  the  long  list  of 
factors  that  swell  the  size  of  normal  unemployment.  There  seems  to  be 
no  theoretical  reason  why  variation  in  total  unemployment  should  be  a 
particularly  early  symptom,  but  in  a  majority  of  cases  it  actually  has 
been.  Covariation  (mostly  negative)  with  all  other  cyclical  series  is,  of 
course,  to  be  expected;  but  we  should  always  keep  in  mind  that  this  must 
not  be  interpreted  to  mean  either  coincidence  of  variations  or  any  con- 
sistent leads  or  lags.  Economic  life  is  not  so  regular  as  that. 

The  formidable  unemployment  of  the  depression  phase  of  the  second 
Kondratieff  shows  up  well,  as  does  the  different  state  of  things  in  the 
prosperity  of  the  third.  But  considering  the  nature  of  the  phenomenon, 


518  BUSINESS  CYCLES 

we  shall  not  be  surprised  to  find  that  it  is  not  so  easy  to  draw  two  lines  of 
different  gradient  through  the  two  Kondratieff  sections  of  the  material 
as  it  is  with  other  series.  The  Kitchins  show  but  little.  This  also  is 
perfectly  natural — observe  the  contrast  with  the  Kitchin  fluctuations  in 
interest — considering  the  importance  of  the  technological  component, 
which  cannot  act  strongly  within  their  span,  and  perhaps  in  part  explains 
why  most  students  who  by  business  cycles  primarily  mean  Kitchins  so 
persistently  divorce  cyclical  from  technological  unemployment.  For  the 
same  reason,  the  violent  movement  of  our  series  in  the  Juglar  rhythm  is 
what  we  should  expect.  For  clusters  of  innovations,  or  "steps  in  evolu- 
tion," assert  themselves  most  obviously  through  Juglars,  as  we  may 
also  verify  by  reference  to  the  behavior  of  the  pig-iron  series.  While  it  is 
interesting  to  note  that,  as  a  rule,  each  of  them  tends  to  absorb  the 
unemployment  of  its  own  creation,  nothing  can,  of  course,  be  inferred 
from  this  about  the  average  duration  of  the  period  during  which  the 
unemployed  individual  is  actually  out  of  work.  This  question  has  been 
investigated  for  the  postwar  period,1  but  it  does  not  concern  us  here. 

With  the  exception  of  the  twin  peaks  in  the  sixties,  which  are  easy 
to  understand,  there  is  one  peak  to  about  every  nine  years.  Unemploy- 
ment above  6  per  cent  occurs  in  16  out  of  64  years,  and  centers,  except  in 
the  cases  of  1862  and  1879,  in  Juglar  depressions.  But  it  is  not  coter- 
minous with  them,  lasting  longer  than  they  did  in  three  cases  (at  the  end 
of  the  sixties,  in  the  eighties,  and  in  the  early  nineties),  while  in  several 
years  which  clearly  belong  to  depression  phases,  that  figure  was  not 
reached.  Except  in  the  eighties,  unemployment  never  went  beyond  8  per 
cent  for  more  than  one  yearly  figure.  This  happened  in  1858  and  1879 
(neglecting  again  1862)  and  was  more  than  made  up  for  in  the  sub- 
sequent year.  Location  of  the  latter  peak,  at  the  threshold  of  prosperity, 
is  particularly  abnormal.  Although  there  is  no  theoretical  reason  to 
expect  even  as  much  regularity  of  amplitude  as  there  is,  it  may  be  safely 
asserted  that  in  any  unemployment  percentage  beyond  6  per  cent  spirals 
and  so  on  counted  for  much  more  than  did  the  direct  effect  of  innovation. 
Troughs,  except  the  one  in  1872,  display  a  habit  of  not  going  beyond,  or 
just  stopping  short  of,  2  per  cent,  which,  even  apart  from  the  historical 
fact  that  they  were  associated  with  overtime,  would,  according  to  our 
guess,  already  spell  supernormal  employment  in  our  (not  merely  the 
statistical)  sense.  The  writer  knows  of  no  evidence,  that  in  the  epoch 

1  See  D.  Weintraub,  Displacement  of  Workers  by  Increase  in  Efficiency  and  Their 
Absorption  by  Industry,  1920-1931,  Journal  of  the  American  Statistical  Association  for 
December  1932.  The  reader's  attention  is  called  to  this  important  piece  of  work,  which 
by  a  much  more  elaborate  method  arrived  at  results  that  are  believed,  if  account  be  taken 
of  the  conditions  of  the  postwar  period,  substantially  to  support  our  conclusions,  in  spite 
of  the  failure  to  connect  displacement  and  absorption  with  the  causation  of  cycles.  We 
shall  return  to  this  in  the  last  two  chapters  of  this  book. 


PHYSICAL  QUANTITIES.    EMPLOYMENT  519 

under  survey  vicarious  unemployment  played  any  major  role.  Corre- 
lation between  unemployment  and  wages  corrected  for  price  level  proves 
nothing.1 

1  For  the  unemployment  problem  of  postwar  times  M.  Jacques  Rueff  investigated  that 
correlation.  On  this,  see  Professor  Pigou's  analysis,  Theory  of  Unemployment,  Chap.  X. 
In  Industrial  Fluctuations,  Part  I,  Chap.  XXI,  he  presents  the  case  against  causal  inter- 
pretation of  such  covariations  so  forcefully  as  to  make  it  superfluous  for  us  to  argue  the 


CHAPTER   X 


Prices  and  Quantities  of  Individual 
Commodities 


A.  Prices  and  Quantities  of  Individual  Commodities  (Including 
Services). — These  should  always  be  studied  together.  Taken  by  itself, 
neither  price  nor  quantity  conveys  its  full  message  or  in  fact  any  that  is 
definite,  and  each  must  be  interpreted  in  the  light  of  its  companion.  In 
general  theory  and  its  statistical  complement,  this  has  of  course  always 
been  recognized.  Students  of  cyclical  variations,  however,  have  some- 
times attempted  to  deal  with  price  variations  alone  or  with  quantity 
variations  alone.1  But  the  true  heroes  of  the  play  and  the  true  variables 
of  the  problem  are  the  price-quantity  pairs  or  points.2  Their  behavior 

1  The  above  is,  however,  not  intended  to  convey  adverse  criticism  of  such  invaluable 
works  as  F.  C.  Mills,  Behavior  of  Prices,  1927;  G.  Tintner,  Prices  in  the  Trade  Cycle,  1935; 
or  A.  F.  Burns,  Production  Trends  in  the  United  States  since  1870,  to  which,  on  the  con- 
trary, the  writer  wishes  to  acknowledge  heavy  obligation,  and  which  he  strongly  recom- 
mends to  the  attention  of  the  reader.     S.  S.  Kuznets'  Secular  Movements  in  Production 
and  Prices,  1930,  is  essentially  a  price-quantity  study  and  so  is  a  book  that  in  some  respects 
may  be  considered  as  a  forerunner  of  much  of  the  later  work  in  the  field  and  keeps,  theoreti- 
cally and  statistically,  a  very  high  level,  although  it  cannot  be  said  to  coordinate  the  theory 
of  its  first  part  with  the  statistical  investigation  that  forms  its  second  part — Marcel  Lenoir, 
Etude  sur  la  Formation  et  le  Mouvement  des  Prix,  1913.     For  price  dispersion,  in  particu- 
lar, compare  M.  Olivier,  Les  Nombres  Indices  de  la  Variation  des  Prix,  1927,  and  O.  Lange, 
Die  Preisdispersion,  1932.     Of  course,  the  whole  of  the  huge  literature  on  price  analysis, 
although  as  yet  almost  unconnected  with  specifically  cyclical  problems  (the  greatest  of  its 
shortcomings  and  the  one  most  urgent  to  remedy)  really  ought  to  be  cited  here.     The 
reader  finds  a  useful  guide  in  the  survey  of  a  number  of  the  most  significant  contributions 
presented  by  L.   O.   Bercaw,   Price  Analysis,  in  Econometrica  for  October  1934.     The 
contributions  of  Henry  Schultz,  M.  Ezekiel,  and  W.  W.  Leontief  should  be  particularly 
mentioned.     See  also  E.  J.  Working,  Demand  Studies  in  Times  of  Rapid  Economic  Change, 
and  H.  Working,  Differential  Price  Behavior  as  a  Subject  for  Price  Analysis,  Econometrica 
for  October  1935.     There  is  also  a  mimeographed  bibliography  compiled  by  the  Depart- 
ment of  Agriculture  group.     C.  F.  Roos,  Dynamic  Economics,  1934,  breaks  new  ground 
in  various  directions.     Warren  and  Pearsons'  book  on  Prices  is  also  relevant  to  the  study 
of  individual  prices.     See  also  the  Cornell  Agricultural  Experiment  Station  Memoir  142. 
This  note  is  of  course  but  a  fragment,  but  it  is  believed  that  it  gives  enough  references  for 
a  start. 

We  shall  not  insist  again  on  the  factual  difficulties  (as  distinguished  from  difficulties 
of  interpretation)  inherent  in  the  material,  which  become  truly  formidable  in  the  case  of 
even  the  simplest  finished  commodity. 

2  Such  a  point  (x,  y]  can  be  represented  by  a  complex  variable,  x  +  yi,  or  as  a  vector 

520 


PRICES  AND  QUANTITIES  OF  INDIVIDUAL  COMMODITIES    521 

is  the  very  heart  of  the  cyclical  mechanism.  Again  a  vast  research 
program  opens  before  us  into  which  we  cannot  enter.  We  must  confine 
ourselves  to  offering  the  following  comments,  which  are  mere  applications 
and  exemplifications  of  the  analysis  of  our  second  and  fourth  chapters. 

1.  We  will  begin  by  recalling  what  has  been  said  on  the  subject  of 
Group  Prices.  Most  of  it  applies  also  to  the  behavior  of  price-quantity 
relations  of  individual  commodities.  Only  it  does  so  with  much  greater 
force  because  individual  peculiarities  must  relatively  increase  in  impor- 
tance as  we  proceed  from  textiles  to  cottons,  from  cottons  to  cotton 
wearing  apparel,  from  this  to  cotton  shirts,  from  cotton  shirts  to  men's 
cotton  shirts,  and  from  these  to  brand  X  of  firm  F.  Again,  we  may  con- 
veniently, though  somewhat  artificially,  class  variations  into  those  which 
are  directly  caused  by  innovation  (and  these  in  turn  into  the  price- 
quantity  variations  of  the  commodities  in  the  production  of  which 
innovation  has  occurred,  innovating  commodities  as  we  call  them,  and 
the  price-quantity  variations,  induced  by  innovation,  in  their  rivals  and 
subsidiaries)  and  those  which  are  simply  responses  to  the  general  business 
situations  that  result  from  the  particular  disequilibria  created  by 
innovation.  It  is  with  the  latter  class  that  such  statistical  uniformity  as 
there  is  must  rest.  And  in  fact  the  impress  of  cyclical  phases  is  in  the 
large  majority  of  cases  clear  enough,  although  not  so  strong  as  students 
expect  who  have  acquired,  from  an  aggregative  view  of  the  cyclical 
process,  a  habit  of  mind  for  which  any  deviation  is  in  the  nature  of  a 
discovery  to  be  wondered  at.  But  every  industry  or  even  concern  has  a 
structural  pattern  of  its  own  and  is  at  every  point  of  time  faced  by  an 
individual!  set  of  environmental  conditions  that  determines  its  reaction. 
Hence  if  we  observe  that  the  price-quantities  of  an  industry  behave 
differently  from  those  of  other  industries,  this  is  no  reason  for  concluding 
that  different  forces  must  be  acting  upon  them  or  that  they  have  (in  any 
but  a  formal  sense)  a  special  cycle  of  their  own,  or  for  denying  the  reality 
of  the  " general"  cycle. 

There  is  much  danger  of  misinterpreting  statistical  findings  in  that 
sense,  especially  if  we  refuse  to  recognize  the  traces  of  the  cyclical  process 
unless  they  display  sufficient  regularity  to  reveal  themselves  to  formal 
methods  of  measurement.  The  reader  should  therefore  realize  from  the 
outset  that  there  is  no  warrant  for  this  attitude  and  that  expectation  from 
our  model  is  not  for  uniformity  but  for  what  we  actually  find,  great 
variety1  of  amplitudes,  periods,  and  sequences  that  does  not  tell  in  the 

since  complex  variables  may  be  geometrically  interpreted  as  vectors  from  the  origin.  Pro- 
fessor W.  W.  Leontief's  important  paper  on  Price-Quantity  Fluctuations  in  the  Business 
Cycle  (Review  of  Economic  Statistics  for  May  1935)  seems  to  the  writer  to  open  up  a  most 
promising  as  well  as  novel  line  of  approach. 

1  "Dispersion  graphs"  on  which  lines  representing  prices  or  quantities  on  a  time  scale 
are  made  to  coincide  for  the  first  year  are  useful  to  inspect  in  order  to  get  a  first  impression 


522  BUSINESS  CYCLES 

least  against  the  presence  of  an  all-pervading  movement  and  does  not  spell 
theoretical,  although  it  does  spell  statistical,  irregularity.  He  should 
train  himself  to  look  upon  every  industry  as  a  resonator  exposed  to  the 
impact  of  a  force  and  responding  to  this  impact  according  to  its  structure. 
These  resonators  are  differently  affected  by  each  individual  impact, 
although  it  is  always  the  same  process  that  impinges  upon  them,  and  they 
would,  because  of  this,  respond  differently,  even  if  they  were  equally 
constructed.  But  they  are  not  and  hence  would  respond  differently 
even  if  equally  affected.  It  follows  that  lead  or  lag  of  individual  price- 
quantity  pairs  as  compared  with  price  level  or  total  output,  as  well  as 
compared  with  other  price-quantity  pairs,  has  but  little  significance  for 
the  causation  and  mechanism  of  the  cyclical  process  per  se. 

£.  We  may  still  schematize,  however.  And  first  we  will  attend  to  the 
behavior  of  the  price-quantity  pairs  of  industries  which,  in  the  sense  of 
the  above  distinction,  are  not  innovating  ones  or  close  relatives  to  these. 
That  distinction  is  not  easy  to  carry  out  in  any  actual  instance  because, 
as  we  know,  innovation  jumps  from  industry  to  industry,  and  intrudes 
into  every  one  of  them  at  some  time  or  other.  But  it  has  some  expository 
virtues.  It  is  convenient,  moreover,  to  start  with  cases  that  approximate 
the  competitive  schema. 

The  impact  of  the  cyclical  sequence  of  business  situations  on  such 
industries  is  by  way  of  enterpreneurs*  expenditure  and  expenditure 
induced  by  it.  Effects  may  be  decomposed  into  nominal  and  real  ones. 
By  the  first  we  mean  effects  on  prices  and  costs  that  cancel  each  other 
and  by  the  second,  effects  that  do  not.  Although  it  would  mean  eliminat- 
ing the  very  essence  of  our  process  if  we  adopted  the  hypothesis  that  that 
expenditure  so  acts  as  to  affect  all  values  in  equal  proportion,1  part  of 
what  actually  happens  does  answer  to  the  pattern  of  such  a  purely  nomi- 
nal alternation  of  "inflation"  and  *' deflation,*'  which  should  induce 
neither  expansion  nor  shrinkage  of  output,  and  in  fact  explains  why  varia- 
tions of  prices  have,  in  all  phases  of  cycles,  much  more  in  common  than 
have  variations  of  quantities.  There  is  some  point,  therefore,  in  correct- 
ing commodity  prices  by  an  index  of  the  level,  though  the  defects  of  these 

of  that  variety.  See,  for  example,  the  charts  on  pp.  12  and  18  of  Professor  Irving  Fisher's 
The  Making  of  Index  Numbers.  The  picture  is,  in  its  pedagogical  value  for  general  pur- 
poses, somewhat  injured  by  the  fact  that  it  covers  the  years  of  the  World  War.  Quantities 
are  seen  to  be  even  more  individualistic  than  prices.  For  a  picture  of  average  rates  of 
change  in  prices  (Professor  Mills  calls  them  "trends")  see  F.  C.  Mills,  Behavior  of  Prices, 
p.  68.  The  graph  has  the  virtue  of  starting  not  too  far  from  a  neighborhood  of  equilibrium 
and  of  illustrating  a  Kondratieff  prosperity.  What  might  be  termed  a  modal  behavior  is 
discernible. 

1  In  that  sense  it  is  of  course  true  that  there  is  no  "one-to-one  relation"  between  any 
single  price  and  the  price  level,  and  that  in  any  multiple  correlation  analysis  price  level 
should,  hence,  be  explicitly  introduced  as  one  of  the  independent  variables. 


PRICES  AND  QUANTITIES  OF  INDIVIDUAL  COMMODITIES    523 

indices  make  it  a  hazardous  operation.     This  has  been  done  by  Mr.  Carl 
Snyder  (Structure  and  Inertia  of  Prices,  quoted  before,  p.  192  for  the 


1840    45     1850    55     I860    65      1870    75      1880    85      1890    95     1900    05     1910  13 
CHART  XXI. — United  States  individual  prices  deflated  (see  Appendix,  p.  1058). 

United  States,  and  p.  194  for  Great  Britain).  He,  however,  seems  to 
claim  for  the  results  an  absence  of  long-time  tendency  and  a  covariation, 
which  not  everyone  will  be  able  to  see.1  Chart  XXI  will  convey  an 
1  However,  a  tendency  to  covariation  may  be  said  to  exist  also  in  the  deflated  figures. 
But  our  version,  both  of  this  finding  and  of  the  covaration  of  undeflated  figures,  may  be 
given  in  the  form  of  an  answer  to  Mr.  Snyder's  question  as  to  what  mainly  determines 
prices:  the  cycle. 


524  BUSINESS  CYCLES 

impression  that  suffices  for  our  purpose.  With  some  obvious  qualifica- 
tions, variations  in  the  sum  total  of  money  wages  would,  under  our  present 
hypothesis,  reflect  those  variations,  both  in  costs  and  receipts,  which 
would  cancel  out  from  the  standpoint  of  the  firms  and  the  wage  earners 
themselves.  In  reality  they  cancel  out  to  some  extent  and,  so  far  as 
they  do,  we  have  no  difficulty  in  taking  account  of  the  cyclical  element  in 
an  ordinary  Marshallian  demand  or  supply  curve  by  introducing,  say, 
the  wage  bill  as  a  parameter  that  will  make  it  shift  in  the  requisite  way.1 
From  the  standpoint  of  some  theories  of  the  cycle,  this  is  in  fact  all  that 
would  be  needed. 

3.  But  as  we  know,  increase  and  decrease  of  expenditure  do  not  act 
in  the  way  just  envisaged  but  so  as  to  produce  shifts  of  "purchasing 
power"  between  households  and  shifts  of  relative  receipts  and  costs 
between  industries  and  even  firms.  There  are,  therefore,  real  changes 
to  which  to  respond  and  this  response  is,  with  but  minor  qualifications, 
uniquely  determined  in  the  competitive  case.  Only,  deterrninateness 
does  not  mean  that  every  industry  and  firm  will  respond  in  the  same 
manner.  It  will  do  so — and  the  analysis  of  the  second  chapter  prepared 
us  for  this — according  to  its  structure  and  its  situation  at  the  moment. 
By  structure  we  do  not  only  mean  the  technological  and  commercial 
setup — including  the  lags  inherent  to  it — and  the  organization  of  the 
industry — such  as  its  relation  to,  and  the  behavior  of,  its  wholesale  and 
retail  trade2 — but  also  the  nature  of  its  products — for  example,  whether 
they  are  subject  to  rapid  change  in  fashion,  readily  variable,  easily  stor- 
able,  or  not — its  financial  habits,  the  support  it  can  expect  from  its  finan- 
cial connections,  finally,  the  mentality,  the  promptness,  and  particularly 
the  horizon,  of  its  managerial  personnel,  its  aversion  to  dismissing  work- 
men, and  the  like.  A  most  important  point  is  the  policy  of  its  firms 
with  regard  to  stocks,  for,  given  the  possibility,  producers  will  in  general 
"speculate"  in  their  own  product.  Data  about  carry-over  of  stocks 
being  exceedingly  scarce  for  prewar  times  (although  valuable  indications 
might  be  unearthed),  we  have  no  choice  but  to  neglect  this  element  of 
cyclical  price-quantity  fluctuations.3 

1  If  we  take  a  linear  demand  function  (entirely  inadmissible  and  even  absurd  though 
it  is)  that  expresses  quantity  (virtually)  sold,  y,  in  function  of  the  price  x  and  the  "catch- 
all"  time,  y  —  a  —  bx  +  ct,  we  may  either  interpret  the  last  item  to  mean  cyclical  varia- 
tion of  the  kind  discussed  above  and  replace  it  by  cw,  w  meaning  wage  bill,  or  we  may  leave 
ct  for  other  "trends"  and  let  a  vary  in  function  of  w,  in  this  case  simply  as  Kw,  K  being  a 
constant.     Of  course  we  can  in  this  case  also  correct  for  price  level  and  then  use  the  original 
equation, 

2  The  very  different  attitudes  of  retailers  toward  their  margins  may  be  cited  as  an 
example.     As  a  rule,  these  margins  are  not  rigid,  but  in  some  branches  they  are. 

8  It  is,  however,  not  too  hazardous  to  draw  conclusions  from  postwar  material.  The 
study  of  R.  H.  Blodgett,  Cyclical  Fluctuations  in  Commodity  Stocks,  1935,  and  various 


PRICES  AND  QUANTITIES  OF  INDIVIDUAL  COMMODITIES    525 

Behavior  of  price-quantity  pairs  also  depends,  at  any  moment  and 
during  every  individual  cycle,  on  the  actual  situation  of  each  industry 
and  firm  at  that  time  and  in  principle  on  all  preceding  situations.  What 
we  mean  by  situation  would  be  very  easy  to  define  if  we  could  start  from 
perfect  equilibrium.  But  this  is  precisely  what  we  cannot  do  now. 
We  must  start  from  the  second  approximation  of  the  fourth  chapter  and 
take  into  account  all  the  underemployment,  all  the  sloppiness,  all  the 
effects  of  indebtedness,  all  the  undigested  leavings  of  previous  phases, 
all  the  imprints  of  chance  events  peculiar  to  the  industry  under  study, 
that  may  be  present  and  account  for  the  diversity  and  the  timing  of  its 
response.  Cycles  may  then  be  skipped  or  there  may  be  what  to  the 
statistician  will  look  like  special  cycles,  although  the  industry  may  be 
reacting  to  nothing  but  "the"  business  cycle  itself.  The  moral  of  the 
story  is  that  only  analysis  of  the  history  and  state  of  an  industry  will 
explain  the  behavior  of  its  price-quantity  pairs.  Unless  research  into 
the  facts  can  proceed  on  these  lines  it  is  useless,  and  we  might  just  as  well 
confine  ourselves  to  stating  our  theoretical  expectations,  for  the  statistical 
graph  will  not  tell  us  more  than  they. 

4.  However,  these  are  only  qualifications,  although  important  ones, 
of  the  broad  truth  that  noninnovating  commodities  which  arc  produced 
either  under  conditions  of  perfect  competition  or  under  conditions  which, 
even  in  the  absence  of  technically  perfect  competition,  enforce  similar 
results  will  display  cycles  in  prices  both  relatively  promptly  and  rela- 
tively strongly  and  hence,  on  the  whole,  tend  to  fluctuate  less  in  quan- 
tity, unless  typically  cyclical  in  nature.  In  order  to  verify  this  we  need 
only  glance  at  the  constituents  of  some  of  the  indices  of  sensitive  prices. 
The  Snider-Persons  index  of  %Z  commodities  (1875-1889),  the  Harvard 
Economic  Society  index  of  13  (from  1923),  and  another  Persons  index 
of  10  commodities  (1890-1922) — for  all  of  them  see  W.  M.  Persons, 
Forecasting  Business  Cycles — list  beans,  barley,  burlap,  coffee,  cotton- 
seed oil  (refined),  cotton  sheetings,  coke,  copper  (ingots),  corn,  corn 
meal,  hides,  hogs,  lard,  oats,  pig  and  bar  iron,  pork,  print  cloth,  rubber,  rye, 
shellac,  silk,  spelter,  steel  scrap,  tallow,  tin,  tin  plates,  tobacco  (leaf),  tur- 

investigations  of  the  Berlin  Institut  filr  Konjunkturforschung  (for  example,  the  article 
on  Lagerhaltung  in  the  Vierteljahrshefte,  1928)  are  particularly  suggestive.  To  the  late 
Professor  Laurits  V.  Birck  of  Copenhagen  is  due  the  idea,  the  present  writer  believes,  of 
making  it  a  criterion  of  cyclical  situations  where  stocks  are  primarily  held — with  producers, 
wholesalers,  retailers,  or  ultimate  buyers.  The  pure  theory  of  the  subject  is  largely  the 
work  of  Professor  Tinbergen.  Mr.  Blodgett's  results  amply  bear  out  the  expectation 
that,  owing  to  the  variety  of  individual  situations  and  of  relevant  considerations,  behavior 
as  to  the  holding  of  stock  must  differ  widely  between  industries  and  firms.  For  instance, 
stocks  in  goods  that  are  perishable  or  subject  to  fashion  will  of  course  be  positively 
associated  with  cycles,  stocks  which  are  due  to  inability  to  sell  and  to  the  difficulty  or 
costliness  of  varying  the  rate  of  production,  negatively. 


526  BUSINESS  CYCLES 

pentine,  wheat,  wood  screws,  wool,  worsted  yarns,  zinc.  Hemp  adorns  a 
corresponding  German  list.  As  the  reader  sees,  the  expectation  of  finding 
those  commodities,  which  according  to  some  theories  are  particularly 
allied  to  cyclical  processes,  is  not  entirely  disappointed.  But  agricultural 
articles,  the  standard  examples  of  almost  perfectly  competitive  condi- 
tions, predominate.  They  would  do  so  still  more  were  it  not  for  two 
circumstances.  First,  though  prices  and  quantities  of  all  of  them  are 
of  course  strongly  influenced  by  chance  variations  of  crops,  some  are 
more  than  others.  Potatoes,  for  instance,  are  dominated  by  this  factor 
much  more  than  wheat  (see  H.  L.  Moore's  highly  satisfactory  demand 
curve  for  potatoes  in  Economic  Cycles,  p.  76,  which  fits  well,  in  spite  of 
the  extreme  simplicity  of  the  assumptions  underlying  its  construction — 
the  price  of  wheat,  for  instance,  shows  much  more  the  influence  also  of 
other  factors) .  Second,  the  purpose  of  deriving  a  sensitive  index  which 
precedes  the  index  of  the  general  price  level,  disqualifies  others  the  prices 
of  which,  like  that  of  milk,  display  the  influence  of  the  cycle  all  right.1 

1  Professor  Mills  accuses  barley,  beans,  corn,  rye,  wheat,  and  many  others  of  irregular- 
ity. This  is,  of  course,  a  correct  formulation  of  the  results  of  applying  his  measurements 
and  his  reference  schema  of  cyclical  movements.  Others  (26  in  all)  he  classes  as  "excep- 
tional," excluding  them  altogether.  But  his  warnings,  op.  cit.,  p.  102,  as  to  the  interpreta- 
tion of  Table  XV  in  his  Appendix  (and  others)  are  but  too  justified.  It  might  be  added 
that  those  results  afford  a  good  illustration  of  our  view  about  such  measurements.  If 
irregularity  of  behavior  be  interpreted  to  mean  (for  example,  p.  553n.)  that  the  prices  of 
commodities  so  designated  "do  not  conform  in  any  systematic  fashion  to  the  cyclical  move- 
ment of  general  prices"  and,  we  suppose,  "exceptional"  behavior  to  mean  irregularity  too 
great  to  be  borne  with,  it  seems  tempting  to  infer  that  the  price-quantity  behavior  of  such 
commodities  is  independent  of  the  cycle  or  contrary  to  the  cyclical  schema.  This,  of  course, 
Professor  Mills  does  not  mean,  and  it  would  be  completely  false.  That  some  of  them  could 
ever  have  been  drafted  for  service  in  an  index  of  sensitive  prices  and  that  they  served  better 
than  pig  iron,  is  in  itself  sufficient  to  refute  it.  Another  argument  is  afforded  by  eggs  and 
butter,  also  classed  as  irregular,  and  poultry,  which  is  voted  exceptional.  Eggs,  butter, 
poultry,  as  well  as  all  the  meats,  classed  as  regular,  beef,  pork,  mutton  (the  writer  is  not 
sure  about  veal)  display  in  their  price-quantity  behavior  as  close  a  relation  to  wage  bill  as 
we  can  expect,  taking  account  of  the  structure  of  these  "resonators."  Again  we  must  for 
this  statement  mainly  rely  on  postwar  data.  The  wholesale  price  of  butter  moves  inversely 
to  Federal  Reserve  Board  Pay  Rolls  in  1927,  but  otherwise  covariation  prevails.  The  gross 
Farm  Value  of  Poultry  and  Eggs,  as  estimated  by  the  Department  of  Agriculture,  goes  well 
with  Pay  Rolls.  Action  of  variation  in  consumers'  expenditure  through  wholesale  prices 
on  farm  prices  is  comparatively  quick  in  all  cases.  See,  for  example,  United  States  Depart- 
ment of  Agriculture,  report  on  hog  situation  to  United  States  Senate,  Feb.  9,  1933,  which, 
though  dealing  with  postwar  fact,  may  still  have  application  to  our  period:  "total  con- 
sumers' expenditure  for  pork  apparently  is  determined  largely  by  the  level  of  consumers' 
incomes.  Consumers'  expenditures  for  a  large  supply  of  pork  are  about  the  same  as  for  a 
small  supply  if  consumers'  incomes  remain  constant."  The  same  is  approximately  true 
of  the  other  articles  mentioned.  To  be  sure,  this  will  result  in  a  variability  of  price  which 
may  look  irregular  in  the  sense  alluded  to,  while,  as  a  matter  of  fact,  it  would  be  difficult 
to  find  much  stronger  instances  of  response  to  cyclical  impulses.  Similar  remarks  apply  to 
some  of  Mr.  Tintner's  measurements.  He  states,  for  instance,  op.  cit.,  that  German  metal 


PRICES  AND  QUANTITIES  OF  INDIVIDUAL  COMMODITIES    527 

The  nonagricultural  articles  call,  in  this  connection,  for  only  two  com- 
ments: cotton  sheeting,  print  cloth,  and  worsted  yarns  are  not  in  the 
least  less  sensitive  than  pig  iron  or  copper.  And  wood  screws  are  par- 
ticularly interesting  as  an  article  the  Marshallian  elasticity  of  demand 
for  which  must  be  extremely  small  and  the  production  of  which  is  carried 
on  under  conditions  certainly  far  removed  from  perfect  competition. 
Professor  Mills  gives  40.2  months  for  the  average  duration  of  its  Kitchin 
(op.  cit.y  p.  545,  No.  291;  bar  iron,  No.  267,  has  39.1). 

This  is  not  to  deny  that  a  more  intensive  study  of  periods,  amplitudes, 
and  relative  phases  (timing)  may  yield  results  that  reveal  cyclical  causa- 
tions. It  is  not  uninteresting  to  note,  for  instance,  that  metal  and  metal- 
product  prices  and  quantities  indicate  Juglars  much  more  clearly  than 
Kitchins  (so,  however,  do  pepper  and  flour,  also  freight  rates1  and  some 
chemicals;  see  Tintner,  op.  cit.),  and  that  foodstuffs  and  raw  and  semi- 
finished textiles  and  textile  machinery  do  the  reverse  (as,  however,  also 
do  other  chemicals  and  timber).  But  no  simple  generalizations  are 
possible  and  no  formal  methods  of  analysis  are  satisfactory.  As  stated 
above,  this  entirely  conforms  to  expectation  from  our  model.  For  the 
study  of  demand  and  supply  it  means  that  Marshallian  curves  fail 
beyond  remedy.  In  specially  favorable  cases  we  may  still  split  the 
actual  movement  into  a  movement,  say,  along  a  demand  curve,  and  a 
movement  of  the  demand  curve.  The  potato  case  cited  affords  an 
instance  in  which  the  first  so  dominates  that,  for  quiet  times,  we  can 
derive  what  looks  like  a  plausible  Marshallian  curve.  Even  there  it 
does  more  to  obscure  than  to  elucidate  the  real  course  of  events.  The 
cases  of  iron,  steel,  copper,  and  other  commodities  of  that  class  afford 
instances  in  which  the  second  so  dominates  that  we  get  II.  L.  Moore's 
positively  inclined  demand  curves,  which,  as  is  universally  recognized 
now  and  as  was  eventually  recognized  by  Moore  himself,  are  no  demand 
curves  at  all,  but  paths  of  cyclical  shift  within  a  family  of  demand  curves, 
each  of  which  is  still  negative  to  the  quantity  axis.  But  to  reason  as  if 
cyclical  variations  of  quantities  supplied  or  demanded  were  movements 
along  any  invariant  curves  is  obviously  inadmissible  and,  as  will  be  seen 
in  later  chapters,  at  the  bottom  of  many  faulty  arguments  about  wages, 
interest  rates,  and  prices  of  capital  goods.  It  is  no  less  inadmissible  to 
assume  that,  while  shifting,  the  forms  of  those  Marshallian  functions 

prices  show  small  response  to  the  cyclical  rhythm.  His  method  yields  this  result.  But  the 
reader  should  look  for  comment  to  the  chart  on  p.  96  in  the  price  study,  quoted  before,  of 
the  Berlin  Institut  ftir  Konjunkturforschung.  Such  procedure  makes  it  still  more  difficult 
to  segregate  the  cases  that  really  present  difficulties  of  interpretation,  such  as  tea,  which  the 
present  writer  entirely  fails  to  understand  (see  Mrs.  Gilboy's  articles  on  Time  Series  and 
the  Derivation  of  Demand  and  Supply  Curves  in  the  Quarterly  Journal  of  Economics  for 
August  1934). 

1  (English)  prices  of  secondhand  ships  move  particularly  well  in  the  Juglar. 


528  BUSINESS  CYCLES 

are  invariant  to  cyclical  phases.  This  is  much  more  important  than  are 
many  of  other  objections  usually  raised  against  the  use  of  "classical" 
analysis  in  this  field. 

B.  Special  Cases. — Recent  work  has  gone  beyond  those  limits. 
Rate  of  change  of  price,1  the  influence  of  past  prices,2  and  a  number  of 
collateral  circumstances  have  been  taken  into  account  (Evans,  Roos, 
and  others.  See  for  a  survey  of  formulae  sufficient  for  our  purpose, 
and  for  a  highly  instructive  application  to  steel,  R.  HH  Whitman,  The 
Statistical  Law  of  Demand  for  a  Producers'  Good,  Econometrica  for 
April  1936).  The  problem  of  the  influence  of  variations  in  real  income 
has  been  analytically  taken  care  of  by  the  Pareto-Slutsky-Schultz 
theory  (see  the  latter's  paper  in  the  Journal  of  Political  Economy  for 
August  1934).  Multiple  correlation  has  lent  its  aid  for  advance  in 
similar  directions  (Ezekiel,  Bean).  Reaction  of  consumers  to  change 
in  price  and  income  has  also  been  the  subject  of  a  League  of  Nations 
investigation  (see  on  part  of  that  project,  Mr.  H.  Staehle's  article  on 
the  behavior  of  immigrants  into  the  United  States,  Econometrica  for 
January  1934).  Some  of  the  results  of  all  those  and  similar  endeavors 
directly  fulfill  our  desiderata.  Others  are  readily  adaptable  to  the 
purposes  of  the  study  of  cyclical  behavior  and  in  fact  may  be  improved 
by  borrowing  from  the  theory  of  cycles.  All  break  down  when  produc- 
tion and  consumption  functions  ("methods"  and  "tastes")  change, 
as  they  unavoidably  do  eventually.  But  for  most  commodities  they, 
particularly  the  latter,  remain  constant  for  considerable  stretches  of 
time.  Any  change  that  occurs  by  way  of  innovation  or  induced  or 
autonomous  change  in  taste  must  be  located  historically.  In  some  cases 
it  shows  in  the  statistical  material,  which,  however,  must  never  be 
relied  on  to  indicate  explanation  unequivocally.  Had  we  nothing  but  a 
historical  graph  of  quantity  of  products  from  the  soil,  we  might  easily 
be  tempted  to  interpret  its  form  in  the  sense  of  decreasing  returns, 
absurd  though  that  would  be. 

1  Professor  Evans  was,  as  far  as  the  writer  knows,  the  first  to  introduce  rate  of  change  of 
price  explicitly  into  the  demand  (and  supply)  functions.     The  simplest  way  of  doing  it  is 
to  add  an  item  of  the  form  hp,  h  being  a  constant  and  p  being  the  time  derivative  of  price. 
See,  for  example,  his  Mathematical  Introduction  to  Economics,  1930,  and  earlier  papers 
This  may  be  interpreted  as  a  way  of  taking  account  of  anticipation.     We  shall  use  this 
idea  later  on  in  another  connection.     That  it  works  with  speculation  in  agricultural  com- 
modities has  been  shown  by  investigations  of  the  Department  of  Agriculture  group. 

2  If  we  treat  quantity  as  a  function  of,  among  other  things,  past  prices — say,  an  average 
of  the  prices  of  the  preceding  five  or  six  years — we  may  mean  either  that  the  capacity 
adjusted  to  those  prices  will  persist  for  a  time  or  that  people  really  intend  to  react  to  those 
past  prices.     Formally,  this  makes  no  difference;  but  it  follows  that,  if  we  observe  that 
farmers  react  in  a  way  which  may  be  amenable  to  interpretation  in  the  latter  sense,  we 
must  not  from  this  alone  infer  that  they  really  mean  to  do  so. 


PRICES  AND  QUANTITIES  OF  INDIVIDUAL  COMMODITIES    529 

We  may  take  another  step  without  leaving  the  precincts  of  our 
present  hypothesis  of  (more  or  less)  perfect  competition,  although  the 
element  of  innovation  will  occasionally  intrude.  Of  the  reasons  why 
fluctuations  of  price-quantity  pairs  of  a  commodity  may  fall  out  of  line, 
in  a  manner  not  explicable  by  the  cyclical  mechanism  without  taking 
account  of  the  facts  of  industrial  structure,  technological  lag  is  one  of  the 
most  important.  To  some  extent  it  is  ubiquitous,  particularly  if  we 
include  adaptation  by  construction  or  reconstruction  of  plant.  But  we 
will  confine  ourselves  to  a  few  outstanding  instances.  Coffee  is  one.1 
We  shall  expect  and  actually  find  in  the  series  of  its  price-quantity  pairs 
traces  of  the  time  shape  of  the  yield  of  coffee  trees.  The  American  price 
— we  deal  with  Brazilian  coffee  and  the  United  States  market  only — varies 
fairly  closely  with  the  Kitchin  phases  (average  duration  for  1890  to  1925, 
according  to  Professor  Mills,  No.  109,  37.0  months),  but  all  the  longer 
movements  are  entirely  out  of  step  moving  under  the  obvious  influence  of 
enormous  waves  of  new  planting.  New  trees  begin  to  bear  after  from 
four  to  seven  years,  after  which  their  yield  varies  with  weather  conditions 
and  the  exhaustion  of  trees  after  bumper  crops.  Such  a  wave  pulled 
the  price  down  to  a  trough  in  the  middle  eighties  and  again,  after  spec- 
tacular rise  to  the  early  nineties,  at  the  beginning  of  this  century. 

Interpretation  is  not  difficult  but  complex.  It  was  a  case  of  innova- 
tion, and  of  an  article  forcing  its  way  into  consumers'  budgets  and 
inducing  a  change  in  taste :  per  capita  consumption  increased  systemati- 
cally to  about  1902,  obviously  not  only  in  consequence  of  increasing 
wealth,  as  we  may  infer  from  a  comparison  of  the  changes  that  occurred 
in  the  relation  between  coffee  and  tea  consumption  in  this  country  and 
England.  Production  being  mainly  for  export,  international  influences 
caused  a  deviation  from  standard  patterns.  But  the  process  was  further 
distorted  by  monetary  disorders  in  Brazil,  which  frequently  put  an 
additional  premium  on  production,  and  by  the  policy  of  government, 
which  subsidized  it  in  various  other  ways  even  before  the  Difesa  do  Cafe 
set  in.  There  is  also  the  influence  of  weather,  of  the  development 
(primarily,  1870  to  1890)  of  steamship  lines,  domestic  transportation 
(until  the  sixties,  coffee  was  transported  from  plantations  to  ports  by 
mules),  the  development  in  the  importing  countries  of  subsidiary  indus- 
tries, and  the  changing  competitive  situation  with  relation  to  the  other 
producing  countries  and  to  tea.  If  we  take  deviations  of  price  corrected 
for  level  and  of  per  capita  consumption  from  curves  drawn  freehand 

1  Of  the  considerable  literature  we  will  mention  only  H.  Roth,  Die  Uebererzeugung  in 
der  Welthandelsware  Kaffee  im  Zeitraum  von  1790-1929,  and  E.  W.  Gilboy's  Study  of 
Coffee  and  Tea,  1850-1930,  previously  cited.  Also  see  Rowe,  Studies  in  the  Artificial 
Control  of  Raw  Material  Supplies,  London  and  Cambridge  Economic  Service,  Special 
Memorandum  35,  1982. 


530  BUSINESS  CYCLES 

through  points  of  inflection  of  smoothed  original  data,  we  find  good 
inverse  covariation  (see  Mrs.  Gilboy's  chart,  op.  cit.9  673),  which  shows 
that  the  "movement  along  a  demand  curve "  was  not  entirely  extinguished 
by  shifts.  But  the  line  through  the  inflection  points  of  the  price  graph 
has  a  shape  all  its  own.  The  corresponding  line  for  per  capita  consump- 
tion since  about  1870  displays  only  what  may  be  called  a  result  trend. 

The  example  displays  very  well  the  list  of  problems  we  encounter  in 
price  analysis  and  particularly,  the  writer  is  grieved  to  say,  the  difficulties 
which  exact  methods  of  analysis  are  sure  to  meet.  Technological  lag, 
by  displacing  and  distorting  effects  and  giving  large  scope  to  miscalcula- 
tion, certainly  will  produce  fluctuations  which  would  be  absent  without  it 
and  cycles  that  are  special  in  this  sense.  But  first,  it  would  be  a  grave 
error  to  assume  that  the  fluctuations  actually  observed  simply  indicate 
the  effects  of  the  presence  of  technological  lag  in  adaptation  and,  second, 
it  would  be  not  less  erroneous  to  think  that  we  have  here  the  case  of  an 
endogenous  fluctuation,  which  of  itself  might  go  on  indefinitely,  possibly 
even  with  increasing  amplitude.  Coffee  simply  responds  to  a  great 
number  of  impulses.  It  creates  cycles  if,  as,  and  when  its  production 
involves  innovation;  it  experiences  the  influence  of  cycles  as  far  as  it 
experiences  cyclical  variations  in  consumers'  expenditure.  The  latter 
effect  is,  if  anything,  more  clearly  marked  than  the  effect  of  the  coffee- 
tree  lag.  And  it  is  to  these  cyclical  and  external  influences  that  its 
fluctuations  are  due.  Only  the  form  of  these  fluctuations  is  shaped  by 
the  structural  properties  of  the  coffee  resonator,  of  which  the  lag  is 
one. 

The  point  we  are  trying  to  make  stands  out  still  more  clearly  in  the 
"cycles  in  animals."  The  case  is  somewhat  simpler  because,  though 
there  are  plenty  of  qualifications,  production  of  animals  for  human 
consumption  is,  for  Germany  and  the  United  States,  primarily  a  domestic 
industry  and,  again  with  exceptions,  one  in  which  the  element  of  innova- 
tion does  not  so  powerfully  disturb  the  picture  as  it  did  in  the  case  of 
coffee.  These  cycles  are  also  held  to  derive  from  the  fact  that  adaptation 
to  any  supernormally  or  subnormally  favorable  situation  is  possible 
only  with  a  technological  lag,  corresponding  to  the  time  it  takes  to  rear 
an  animal  to  the  requisite  age,  substantially  fixed1  and  the  same  for  all 

1  As  the  reader  sees,  we  are  simplifying  to  the  utmost.  None  of  the  above  statements 
is  entirely  true.  To  begin  with,  even  if  the  period  of  rearing  and  fattening  were  exactly 
equal  for  all  the  farmers  of  a  country,  the  moment  when  they  get  ready  with  their  supply 
would  differ  somewhat  because  in  some  cases  and  within  certain  limits  the  decision,  for 
instance,  to  rear  more  hogs  can  be  immediately  given  effect  to,  while,  in  other  cases  and 
beyond  certain  limits,  time-consuming  rearrangements  on  the  farm  may  be  necessary. 
Second,  the  situations  facing  farmers,  particularly  in  different  parts  of  a  country,  are  not 
equal.  The  effect  of  a  favorable  or  an  unfavorable  fodder-meat  ratio  depends  on  the 
possibilities  in  other  lines  of  production.  Third,  there  is,  both  for  the  initial  decision  and 


PRICES  AND  QUANTITIES  OF  INDIVIDUAL  COMMODITIES    531 

producers  who,  being  faced  by  the  same  situation,  have  to  make  their 
decisions  at  about  the  same  time.  It  should  be  observed  at  once  that 
it  is  further  necessary  to  assume  that  they  will  take  no  account  of  their 
competitors'  actions  until  they  all  of  them  come  out  with  their  product. 
Obviously,  this  assumption  is  unjustified,  for  the  behavior  of  competitors 
is  no  secret.  Therefore,  waves  started  by  a  single  disturbance  would, 
notwithstanding  the  lag,  soon  die  down.  The  idea  that  one  disturbance, 
once  having  caused  an  exceptionally  favorable  or  unfavorable  situa- 
tion, will  thereby  create  a  wave  that  might,  under  its  own  steam  as  it 
were,  go  on  forever  is  manifestly  absurd.  Producers  may  and  actually 
do  react  in  those  cases  in  such  a  way  that  the  mass  effect  of  their  adaptive 
action  will  create  a  disequilibrium  in  the  direction  opposite  to  the  devia- 
tion to  which  they  intended  to  adapt  themselves;  and  rebound  from  the 
untenable  position  is  similarly  likely  to  outrun  the  goal.  But  eventually 
they  would  learn  the  lesson.  It  is  nothing  short  of  thoughtless  to  trust 
the  constants  of  some  equation  to  supply  proof  of  the  contrary.  We 
recall  what  has  been  said  on  the  subject  in  the  second  and  fourth  chapters 
and  conclude  that,  since  those  waves  go  on  apparently  indefinitely  and 
since  they  are  (comparatively)  so  regular,  they  can  precisely  not  be 
endogenous  in  the  full  sense  of  the  word,  but  must  be  kept  going  by 
shocks  from  without  or  by  a  generating  mechanism.  This  mechanism 
can  be  only  the  cycle — particularly,  the  Kitchin,  but  also  the  Juglar — 
acting  through  consumers'  expenditure. 

That  this  is  so  can  be  readily  seen,  for  instance,  in  the  standard  case, 
the  hog  cycle.1  Hog  prices,  on  the  whole,  move  well  with  all  three  cycles 

later  on,  room  for  choice  and  also  for  revision  of  choices  made.  Even  if  all  had  actually 
produced  the  same  type  of  animal  with  a  view  to  having  it  ready  at  about  the  same  date, 
necessity  for  selling  at  that  date  differs  widely.  But  although  these  and  similar  considera- 
tions should  make  us  less  confident  in  the  interpretation  of  certain  striking  regularities, 
we  may  take  it  that,  for  the  limited  purposes  we  now  have  in  view,  they  are  of  secondary 
importance 

1  But  it  is  the  same  with  cattle  and  lambs.  All  those  cases  are  complicated  by  inter- 
commodity  relations  and  jointness  in  animal  products  and,  of  course,  such  external  events 
as  weather  (including  its  effects  on  the  price  of  fodder),  variations  in  consumers'  tastes,  and 
so  on.  Sheep  production  is  nevertheless,  in  this  country  at  least,  a  very  good  case,  in  spite 
of  the  complicated  relation  of  the  consumption  of  lamb  and  mutton  to  that  of  beef,  pork, 
poultry,  and  veal  (to  the  last  of  which  it  shows  some  inverse  relation) .  The  sheep  popula- 
tion in  this  country  did  not  display  any  marked  tendency  from  1890  to  1913,  but  three  very 
regular  cycles.  Recovering  from  a  drop  it  rose  to  a  maximum  in  1893,  dropped  to  a 
minimum  in  1897,  reached  an  almost  equal  maximum  in  1902,  and  a  somewhat  higher  one 
in  1909,  after  which  it  dropped  sharply.  Farm  value  of  sheep  per  head  has  maxima  in 
1893,  in  or  about  1900  and  1906,  Chicago  price  (medium  to  choice)  in  1892,  1898,  and  1905. 
Sheep  price  precedes  sheep  population  by  about  two  to  three  years.  There  was,  of  course, 
the  transition  from  production  for  wool  to  production  for  meat  and  the  increasing  emphasis 
on  lamb  production.  Those  six-,  eight-,  or  nine-year  cycles  cannot  be  simply  linked  to  the 
lag;  on  the  contrary,  they  point  to  other  factors.  They  seem  to  have  much  more  to  do  with 


532  BUSINESS  CYCLES 

(see,  for  example  for  Germany,  the  chart  on  p.  91  of  the  new  price  study, 
previously  quoted,  of  the  Berlin  Institute,  covering  by  two  scries  pork 
till  1863,  hogs  since — nearly  150  years.  For  this  country  and  over  his 
period,  Professor  Mills  gives  a  good  mark  to  his  two  hog  entries,  Nos. 
15  and  16,  duration  of  cycle,  38.8  and  38.4  months,  respectively).  Their 
relation  to  cyclical  changes  in  consumers'  income  is  too  obvious  to 
require  proof.1  It  is  true  that  these  shifts  of  demand  are  combined 
with  strong  movements  "along  demand  curves."  Receipts  of  hogs  at 
markets  or,  for  that  matter,  slaughter,  are  related  inversely  to  prices  of 
hogs  (see,  for  example,  Mr.  Hanau's  graph  for  Prussia,  1900-1913,  p.  18 
of  the  first  edition  of  his  study)  and  positively  to  the  relation  between  the 
prices  of  hogs  (or  pork)  and  of  fodder  that  prevailed  about  18  months2 
earlier.  Prices  of  hogs  could  almost  be  forecast  from  this  relation  alone, 
and  it  therefore  seems  at  first  sight  that  that  market  is  entirely  dominated 

the  sweep  of  the  Juglar,  although  the  relation  is  as  blurred  as  we  might  expect  it  to  be. 
Mr.  M.  Ezekiel's  most  interesting  study,  Factors  Related  to  Lamb  Prices,  Journal  of  Politi- 
cal Economy  for  April  1927,  entirely  free  from  special-cycle  obsessions,  and  a  landmark  for 
that  kind  of  work,  finds  a  68.3  per  cent  " determination"  of  the  price  of  dressed  lamb  (in 
the  multiple  correlation  sense)  by  Mr.  Snyder's  General  Price  Level.  This  is  quite  enough 
to  establish  our  point,  although  correlation  with  a  pay-roll  index  would  have  been  more  rele- 
vant. That  business  activity  figures  among  determining  factors  only  with  0.7  per  cent 
does  not  prove  anything  against  us,  because  it  is  measured  by  the  Harvard  Price  Index  of 
Business  Cycles,  deflated  by  the  B.  L.  S.  wholesale  price  index.  Mr.  Ezekiel  himself 
expresses  doubts  on  the  subject  in  his  note,  op.  cit.,  p.  248. 

We  cannot  and,  for  our  purpose,  need  not  go  into  the  ever-expanding  work  on  hogs. 
We  merely  quote  Haas  and  Ezekiel,  Factors  affecting  the  Price  of  Hogs,  Department  of 
Agriculture  Bulletin  1440,  November  1926;  (S.  Benner,  the  discoverer  of  the  hog  cycle, 
as  far  as  the  writer  knows,  should  not  be  forgotten  however:  Prophecies  of  Future  Ups  and 
Downs  in  Prices,  1876);  Sarle,  Forecasting  the  Price  of  Hogs,  American  Economic  Review, 
1925;  Sewell  Wright,  Corn  and  Hog  Correlations,  Department  of  Agriculture  Bulletin  1300; 
A.  E.  Taylor,  Corn  and  Hog  Surplus  of  the  Corn  Belt,  1932;  and  A.  Hanau,  Die  Prognose 
der  Schweinepreise,  Vierteljahrshefte  zur  Konjunkturforschung,  Sonderhcft  2,  revised  ibid. 
18.  For  cattle,  see  the  same  author's  study,  ibid.,  Sonderheft  13.  Grateful  acknowledg- 
ment should  be  made  for  helpful  suggestions  derived  in  this  whole  range  of  subject  from 
L.  H.  Bean,  The  Farmer's  Response  to  Price,  Journal  of  Farm  Economics,  1929. 

1  See  O.  V.  Wells,  Farmers'  Response  to  Price  in  Hog  Production  and  Marketing,  Depart- 
ment of  Agriculture  Technical  Bulletin  359,  April,  1933,  in  which  study  this  element  has  been 
considered,  p.  8.     Notice  also  the  behavior  of  commercial  slaughter,  p.  37.     Haas  and 
Ezekiel,  op.  cit.,  arrive  again  at  a  very  low  estimate  of  the  influence  of  the  cycle  by  taking, 
this  time,  not  only  the  Harvard  Price  Index,  but  also  an  index  of  prices  of  industrial  stocks 
as  a  series  to  correlate  with,  for  which  there  is  still  less  justification.     The  graph  on  p.  26 
of  that  same  study  shows  the  Kitchin  almost  ideally. 

2  There  was  some  "technological  change"  in  fattening  which  seems  to  have  affected  the 
period,  but  it  was  about  that  length  in  the  last  prewar  decade  both  in  Germany  and  in  the 
United  States.     It  is,  however,  not  quite  convincing,  for  the  period  of  gestation  plus  the 
period  of  rearing  and  fattening  is  more  nearly  15  months.     That  the  farmer  does  not  act 
at  once  upon  a  given  hog-fodder  price  relation  is  plausible.     It  is  less  plausible  that  it  should 
always  take  him  just  3  months  to  arrive  at  a  decision. 


PRICES  AND  QUANTITIES  OF  INDIVIDUAL,  COMMODITIES    533 

by  the  variations  of  supply,  which  in  turn  are  entirely  mechanical.  The 
inference  about  the  presence  of  a  special  cycle  unrelated  to  the  business 
cycle  which  some  students  draw  is,  nevertheless,  entirely  unwarranted. 
What  we  behold  when  looking  at  those  hog  graphs  that  are  so  remarkably 
regular,  is  nothing  but  the — wavelike,  to  be  sure — working  of  a  particular 
apparatus  of  response. 

C.  The  Cycle  in  Shipbuilding. — This  cycle,  made  famous  by  Professor 
Tinbergen,1  serves  to  illustrate  a  lag  phenomenon  incident  to  all  time- 
consuming  construction  of  plant  and  equipment  and  therefore  differs 
(also  in  other  respects)  materially  from  the  hog  case.  Our  discussion 
continues  an  argument  opened  in  the  fourth  chapter,  sec.  E.  Let  us 
consider  total  tonnage  as  a  certain  function  of  time  (cyclical  time  in  the 
first  instance,  but  later  to  be  linked  with  historical  time),  say,  /(£);  and 
let  us  identify  rate  of  change  in  total  tonnage  in  a  first  approximation 
with  shipbuilding,  which  therefore  is  f'(t).  If,  at  any  point  of  time,  total 
tonnage  be  above  normal — whatever  that  may  mean — this  will  send  down 
freight  rates  and  reduce  shipbuilding  (it  cannot  make  the  increment  of 
tonnage  negative).  If  total  tonnage  be  below  normal,  the  reverse  will 
happen  and  after  a  period  #,  determined  by  the  time  it  takes  for  shipping 
companies  to  order  and  for  shipbuilders  to  build  new  tonnage,  total 
tonnage  will  increase  according  to  the  intensity  with  which  carriers  react. 
This  intensity  is  assumed  to  be  a  constant  a,  and  is  measured  in  terms  of 
the  increase  or  decrease  in  tonnage  ordered  that  corresponds  to  some  unit 
deviation  of  tonnage  from  normal.  Shipbuilding  or  rate  of  change  in 
tonnage  is  thus  linked  to  that  tonnage  which  existed  at  time  t  —  $, 
hence  f'(t)  =  ~af(t  —  $).  This  functional  equation  is  treated  in  the 
familiar  way,  i.e.,  by  substituting. 

f(t)  =  eat+P  =  Ceat 

Clearing  for  Ceat,  we  get  a  =  —  ae~a*>  and  if  we  put  —  <*$  =  x  +  iy>  we 
get  an  exponential  with  a  complex  exponent  the  imaginary  part  of  which 
will  give  us  periodic  fluctuations.2  These  fluctuations  in  tonnage  can 

1  Ein  Schiffbauzyklus?     Wdhvirtschaftliches  Archiv  for  July  1931.     What  will  be  said  in 
the  text  should  not  be  construed  as  adverse  criticism  of  Tinbergen's  work.     In  later  papers 
he  has  shown,  in  particular,  that  he  is  fully  aware  of  those  elements  of  the  case  that  we  are 
going  to  stress. 

2  There  is  some  danger  of  the  layman's  misunderstanding  this  procedure.     The  complex 
exponent  is,  of  course,  chosen  because  we  know  that  there  are  fluctuations  in  the  phenom- 
enon.    But  that  these  fluctuations  are  due  to  the  lag  that  enters  into  the  functional,  from 
which  they  are  then  by  means  of  that  complex  exponent  deduced,  is  really  a  new  hypothesis, 
which,  to  be  sure,  may  derive  justification  from  good  fit  of  results — we  do  not  believe  it 
does — but  which  must  not  be  mistaken  for  a  result  following  from  the  original  setup  itself. 
This  could  equally  well  be  satisfied  with  an  aperiodic  solution,  which  the  existence  of  the 
fluctuations  would  not  prove  to  be  wrong.     They  would  simply  have  to  be  ascribed  to 
something  else. 


534  BUSINESS  CYCLES 

then  be  represented  by  a  composite  of  cyclical  and  of  aperiodic  move- 
ments, periods  depending  on  &  and  a  only.  The  composite,  cleared  of 
meaningless  solutions,  is  extremely  pliable,  the  introduction  of  this  piece  of 
apparatus  a  matter  of  congratulation. 

But,  as  applied  to  our  subject,  this  chain  of  reasoning  gives  rise  to 
various  doubts.  To  begin  with,  fluctuations  of  this  kind  cannot  be  called 
endogenous  in  the  sense  of  being  self-generating,  for  they  obviously 
depend  for  their  existence  on  some  disturbance  that  starts  them.  Second, 
they  can  be  self-perpetuating  only  if  we  assume  a  very  peculiar  form  of 
reaction  by  carriers,  which  violates  not  only  the  general  assumptions  of 
economic  theory  about  rationality  of  behavior  but  is  at  variance  also 
with  commonsense.  We  cannot  reasonably  assume  that  reaction  to, 
say,  abnormally  favorable  freight  rates  will  be  mechanical  and  proceed 
without  any  consideration  of  the  causes  and  probable  duration  of  that 
state  of  things  and  of  the  effects  of  simultaneous  action  by  the  whole 
trade.  It  is  no  valid  objection  that  results  will  only  show  with  a  lag  and 
impinge  upon  a  situation  that  has  changed  meantime,  so  that  even 
reaction  correct  ex  visu  of  the  moment  in  which  it  was  decided  on  may 
yet  be  proved  by  the  outcome  to  be  wrong :  for  this  would  not  happen  if 
no  other  process  were  at  work  or  if  no  further  external  disturbance 
occurred — in  which  cases  it  is  that  process  or  this  disturbance  and  not 
any  automatism  inherent  to  shipbuilding  that  causes  the  discrepancy  and 
further  fluctuations.  Third,  the  argument  involves  neglect  of  the 
regulating  influence  of  a  reaction  of  prices  of  new  ready  ships,  second-hand 
ships,  and  newly  ordered  ships.  Fourth,  there  is  the  objection  to  using  a 
trend  line  as  a  normal  from  which  to  measure  deviations  of  tonnage. 
Since  this  trend  itself  is  the  work  of  growth  and  cycles,  its  elimination 
removes  part  of  the  essence  of  the  process.  Nor  can  it,  fifth,  be  taken  as 
a  standard  by  which  to  judge  correctness  of  reaction,  for  cyclical  varia- 
tions of  demand  must  be  taken  into  account  both  by  carriers  and  by 
the  economist  who  observes  their  behavior:  freight  rates  are  not  an 
invariant  function  of  tonnage  alone  and,  as  soon  as  this  is  recognized, 
there  is  an  end  of  this  particular  cycle. 

Inspecting  Chart  XXII,  we  find  the  implications  of  this  fully  borne 
out  by  the  facts  of  the  case.  Of  course  we  see  the  effects  of  all  the 
peculiarities  of  the  industry — the  effects  of  subsidies  and  policies  of 
prestige  embarked  upon  by  other  nations  but  influencing  the  British 
situation;  of  the  fact  that  shipping  and  ship-building  reflect  economic 
and  political  conditions  all  over  the  world;  of  the  relation  that  trans- 
atlantic freight  carrying  bears  to  emigration;  of  the  fact  that  shipping  will 
share,  to  a  large  extent,  the  fate  of  international  commodity  trade,  which 
itself  deviates,  especially  for  longer  periods,  from  the  contour  of  the 
cycle;  and  so  on.  Above  everything,  we  must  attend  to  the  fact  that, 


PRICES  AND  QUANTITIES  OF  INDIVIDUAL  COMMODITIES    535 

technologically  and  commercially,  important  innovations  took  place  in 
shipping  and  shipbuilding,  which  will  account  for  the  long-time  tendency 
of  freight  rates  to  fall  even  after  the  general  rise  in  prices  had  set  in,  and 
partly  also  for  rate  wars.  Considering  all  this,  cycles  show  well  enough 
both  in  rates  and,  particularly,  in  construction.  Their  movements, 


ANNUAL  :. 
,'  INCREASE: 
:  IN  TOTAL: 

TONNAGE ; 


PRICE  "OF  NEW, 
READY  CARGO 


1870     1875      1880      1885     1890     1895      1900     1905     1910      1915       1920      1925      1930    1934 
CHART  XXII. — British  shipping  and  shipbuilding  (see  Appendix,  p.  1059). 

inverse  to  each  other  in  a  long-time  sense  for  the  period  covered  by  the 
chart,  and  substantially  synchronous  and  in  the  same  direction  in  a 
short-time  sense,  are  exactly  what  we  should  expect  in  the  case  of  the 
cyclical  process  acting  on  this  particular  resonator  and  can  be  fully 
accounted  for  without  the  element  of  lag.  The  writer  believes,  in  fact, 
that  the  latter  had  very  little  influence  and  that  such  verification  of  the 
opposite  view  as  may  be  derived  by  formal  methods  is  largely  delusive. 
D.  Entrepreneurial  Price  Policies. — If  the  entrepreneurial  act  con- 
sists in  the  production  by  a  new  method  of  a  commodity  already  produced 
under  conditions  of  perfect  competition  which  the  entrepreneur  is  powerless 


536  BUSINESS  CYCLES 

to  alter,  so  that  he  is  confronted  by  an  individual  demand  curve  of 
infinite  elasticity,  the  price-quantity  pairs  of  the  industry  in  which  the 
innovation  occurs  and  of  its  subsidiaries  and  competitors  will  be  affected 
in  ways  that  do  not  call  for  additional  comment.  Occurrence  of  innova- 
tion must,  as  mentioned  above,  be  located  historically1  in  every  indi- 
vidual case,  a  task  necessarily  preliminary  to  any  thorough  price-quantity 
analysis.  Recalling  what  has  been  said  before  and  using  traditional 
concepts,  we  may  now  define  the  ultimate  task  of  such  analysis  as  follows : 
given  time  series  of  actual  price-quantity  behavior,  to  resolve  them  into 
five  components,  namely,  movements  along  supply  or  demand  curves, 
shifts  (including  changes  in  form)  of  demand  curves  within  invariant 
indifference  varieties  (constant  tastes),  shifts  (also  including  changes  of 
form)  of  supply  curves  within  invariant  production  functions,  autono- 
mous changes  of  tastes,  changes  of  production  functions  (innovation). 
The  task  is  difficult  if  numerical  exactness  is  the  goal.  It  is,  though 
laborious,  not  difficult  if  we  content  ourselves  with  a  rough  common- 
sense  approach.  In  particular,  there  is  no  difficulty  in  accounting  for 
the  behavior  of  price-quantities  in  the  prolonged  periods  of  adjustment 
that  are  sometimes  induced  by  innovation  and  during  which,  many  or 
most  firms  continuing  to  produce  at  a  loss  for  years,  the  survival  of  the 
unfit  creates  that  species  of  overproduction  with  which  we  are  familiar. 
Other  deviations  of  price-quantity  pairs  of  innovating  industries  from 
the  average  of  all  commodities,  which  might  easily  be  mistaken  for  spe- 
cial cycles  or  trends,  are  not  less  easy  to  understand. 

But  if  innovation  consists  in  the  introduction  of  a  new  commodity, 
the  entrepreneur  finds  himself,  as  we  have  seen,  almost  invariably  in  an 
imperfectly  competitive  situation.  In  most  cases  his  enterprise  also 
impinges  on  a  sector  in  which  imperfection  prevails  independently  of  it, 
so  that  the  little  which,  for  our  purpose,  it  is  necessary  to  add  to  what  has 
been  said  on  the  subject  in  the  second  and  fourth  and  in  the  historical 
chapters,  can  be  conveniently  dealt  with  under  this  aspect.  We  then 
meet  again  the  phenomenon  which  we  call  stability  of  price  if  we  approve, 
and  rigidity  of  price  if  we  disapprove  of  it.  Although  there  is  little 
reason  to  believe  that  this  stability  or  rigidity  has  been  on  the  increase 
during  the  last  fifty  years,  and  that  it  really  merits  the  attention  which  has 
been  paid  to  it  of  late,  it  is  important  for  our  purpose  to  understand  what 
it  is  that  prevents  so  many  prices  of  new  as  well  as  of  old  commodities 
from  participating  in  the  cyclical  fluctuations  as  prices  under  perfectly 
competitive  conditions  would. 

1  It  follows,  however,  from  our  discussion  in  the  third  chapter  that  the  occurrence  of 
innovation  may  be  indicated  by  the  behavior  of  marginal  cost  corrected  for  change  in 
prices  of  factors.  The  development  of  this  tecbftique  and  corresponding  fact  finding  are 
urgent  desiderata. 


PRICES  AND  QUANTITIES  OF  INDIVIDUAL  COMMODITIES    537 

Traditionally  sticky  and  authoritatively  regulated  prices  do  not, 
from  our  standpoint,  differ  very  much.  That  cycles,  Kitchins  in  partic- 
ular, will  in  many  cases  be  missed  by  commodities  and  services  that  fall 
in  either  class  is  too  obvious  to  detain  us.  But  three  points  deserve 
notice — all  of  them  relevant  also  to  the  examples  that  follow.  First, 
any  industry  or  firm  which  asks  a  public  service  commission  to  sanction 
an  increase  or  a  decrease  in  price,  must  know  that  the  first  is  extremely 
difficult  to  grant  and  that  the  latter  is,  for  the  same  reason,  likely  to  be 
definitive.  It  requires  spectacular  emergencies — inflation  may  not  be 
sufficient — to  "justify"  an  increase  in  the  eyes  of  an  invariably  hostile 
public  opinion.  This  is  an  external  factor.  Second,  it  does  not  neces- 
sarily follow  that  quantities  will  now  fluctuate  more  than  they  would  with 
a  more  variable  price.  As  it  happens,  many  of  the  commodities  in 
question  meet  with  a  demand  of  low  elasticity,  which  also  does  not  much 
shift  in  the  course  of  cycles.  It  is  often  doubtful  whether  quantity 
sold  would,  in  an  ordinary  depression,  be  appreciably  greater  if  prices 
were  promptly  reduced.  Third,  in  not  unimportant  cases  demand, 
without  being  inelastic,  is  such  as  to  react  only  to  price  reductions  so 
considerable  as  to  be  normally  out  of  the  question  without  an  important 
innovation.  Quite  frequently,  reduction  will  have  sense  only  if  the 
intention  is  to  tap  new  strata  of  consumers.  But  such  a  possibility  is  not 
necessarily  associated  with  any  particular  phase  of  the  cycle,  and  may  well 
offer  itself  in  prosperity. 

We  will  next  consider  the  case  of  a  few-firm  industry  the  units  of 
which  have,  through  mere  sluggishness  of  business  spirit,  for  some  time 
been  left  in  undisturbed  possession  of  the  field.  The  firms  are  assumed  to 
be  no  less  sluggish  themselves  and  to  have  settled  down,  but  without 
agreements,  in  an  oligopolistic  quasi-equilibrium  buttressed  by  product 
differentiation.  Such  an  oasis,  it  is  true,  is  not  easy  to  find — most 
instances  that  one  might  think  of  proving  to  be  spurious  on  analysis — 
but  it  will  serve  to  illustrate  a  point  we  wish  to  make.  Let  prosperity 
set  in.  The  firms  will  enjoy  brisker  sales,  but  unless  they  are  unable  to 
increase  their  output  without  incurring  higher  cost  per  unit,  they  will 
not  readily  take  advantage  of  the  situation  by  raising  their  prices;  for  in 
this  case  price  cannot  impersonally  rise  as  does  the  price  of  wheat. 
Somebody  who  can  be  identified  has  got  to  do  the  thing;  and  he  risks 
losing  custom,  even  if  the  others  eventually  follow,  which  at  least  some 
of  them  may  not  do  at  all,  because  they  hope  to  conquer  some  ground 
that  they  did  not  dare  to  invade  initiatively.  Similar  considerations 
apply  on  the  downgrade.  It  should  be  observed  that  the  very  stable 
prices  that  statistics  will  report  while  such  a  state  of  things  lasts,  are  not 
inflexible  in  the  same  sense  in  which  a  traditional  or  regulated  or  agreed 
price  is.  They  could  vary  at  any  moment  and  are  chosen  on  strictly 


538  BUSINESS  CYCLES 

rational  considerations,  with  a  view  to  what  under  the  circumstances  is 
maximum  advantage  over  time.  We  have  in  this  case  assumed  sluggish- 
ness, not  in  order  to  suggest  irrationality,  but  only  in  order  to  exclude 
the  spirit  of  innovation  and  propensity  to  fight. 

Irrational  elements  of  course  come  in  to  help  to  petrify.  In  spite  of 
both  rational  and  irrational  elements,  the  situation  is  not  likely  to  last 
in  a  world  pervaded  by  our  process.  But  if  existing  firms  are  satisfied 
with  it,  they  may  and  often  do  attempt  to  peg  it — which,  for  some  of  them, 
implies  readiness  to  forego  such  individual  opportunities  as  they  may 
have  reason  to  expect  from  the  incessant  revolutions  wrought  by  that 
process — and  to  adapt  it  by  corporative  action  to  changing  conditions. 
They  often  try  to  "rationalize"  it  by  agreeing  to  outlaw  "cross  hauling" 
— which,  however,  is  not  easy  to  recognize  in  the  case  of  product  differen- 
tiation. They  may  get  together  in  order  to  "educate"  each  other  up  to 
an  "ethical  code."1  Or  they  may  simply  decide  to  prevent  their  resale 
agencies  from  starting  on  a  warpath  of  their  own — one  of  the  motives, 
though  not  the  most  important  one,  for  the  movement  toward  Resale 
Price  Maintenance.2  They  also  may  corporatively  adopt  a  policy  of 
Basing  Point  or  of  Delivered  Prices,  as,  for  example,  has  been  done  in 
this  country  by  the  steel,  cement,  lumber,  paper,  petroleum,  beer,  rubber, 
glass,  fertilizer,  flour,  and  sugar  industries.  We  are  not  here  concerned 
with  the  merits  and  demerits  of  these  practices,  into  the  discussion  of 
which  enters  so  much  "theory"  that  really  deserves  the  quotes.  But  it 
should  be  observed  that,  while  some  of  them  may  work  in  the  direction 
of  perfect  competition  and  none  of  them  necessarily  spells  rigidity  of  price, 
yet  all  of  them  require  machinery  which  is  difficult  and  sometimes  costly 
to  set  into  motion.  We  had  examples  in  the  historical  chapters.  Indus- 
trial resonators  are  structurally  affected  by  those  policies,  and  failure  of 
prices  to  react  to  cyclical  situations  is,  in  a  very  obvious  way,  often 
accounted  for  thereby. 

If  policies  of  that  type  issue  in  a  cartel  of  the  German  kind,  the  work- 
ing of  the  brittle  monopoly  that  may  be  enjoyed  by  the  industry  in  the 
intervals  between  recurrent  breakdowns  reveals,  first,  the  main  rational 
and  purely  economic  reason  for  price  stability  in  a  monopoly  situation. 

1  Such  attempts  are  sometimes  unintentionally  humorous.     But  their  meaning  is  not 
disposed  of  by  the  proof  that  some  arguments  used — such  as  the  argument  that  total 
average  cost  must  always  be  covered — are  erroneous.     See,  for  an  instance,  E.  C.  Brown, 
Price  Competition  in  the  Commercial  Printing  Industry  of  Chicago,  Journal  of  Political 
Economy,  1930.     Notice,  in  particular,  the  Code  of  Ethics  adopted  by  the  United  Typoth- 
etae  of  America,  p.  199.    How  many  of  us,  and  for  how  varied  reasons,  yearn  for  the 
spirit  of  the  Middle  Ages! 

2  Examples  abound;  see,  for  instance,  E.  T.  Goether,  Resale  Price  Maintenance  in  Great 
Britain,  Quarterly  Journal  of  Economics  for  August  1934.     This  paper  has  the  merit  of 
bringing  out  very  clearly  that  such  attempts  are  like  canoes  trying  to  live  in  a  stormy  sea. 


PRICES  AND  QUANTITIES  OF  INDIVIDUAL  COMMODITIES    539 

It  has  been  mentioned  above  in  another  connection:  a  monopolist,  if  he 
faces  a  demand  which,  within  the  useful  interval,  either  is  always  insensitive 
to  price  or  becomes  so  when  shrinking  in  depression,  will  have  no  motive 
to  reduce  price,  while  in  perfect  competition  it  would  always  fall.  This 
is  so  with  many  highly  finished  consumers'  articles  of  small  importance 
in  the  household  budgets  and  with  some  important  producers'  goods  that 
individually  contribute  but  little  to  the  total  cost  of  any  product.  But 
it  should  again  be  emphasized  that  the  departure  of  such  prices  from 
the  cyclical  pattern  which  would  prevail  under  conditions  of  perfect  com- 
petition does  not,  in  the  case  envisaged,  imply  corresponding  deviation  in 
quantities,  which  do  shrink,  no  doubt,  but  not  primarily  because  prices 
stay  up  and  not  much  more  than  they  would  in  the  absence  of  monopoly. 
As  far  as  this  goes,  therefore,  certain  well-known  propositions  about 
price  rigidities  that  intensify  depression  do  not  follow,  and  there  is  some 
justification  for  that  attitude  taken  both  by  business  interests  and 
politicians  which  in  our  days  crystallized  into  the  NRA  legislation.1 

Second,  that  tendency  of  monopoloid  situations  to  produce  stable 
prices  is  reinforced,  even  if  there  is  no  cartel  or  monopoly  in  a  strict  sense, 
by  the  practice  of  providing  capacity  for  cyclical  peak  demands .  As  stated 
before  (fourth  chapter),  this  practice  may  result  in  the  prices  of  the  most 
cyclical  industries  becoming  the  least  cyclically  variable  of  all.  This 
phenomenon  would  also  be  impossible  in  perfect  competition.  It  does 
intensify  fluctuations  in  quantities  but  rather  mitigates  than  intensifies 
cyclical  difficulties.  Third,  the  rational  but  extraeconomic  considerations 
that  have  been  mentioned  above  greatly  influence  a  capitalistic  combine 
which  knows  that  it  is  unpopular,  that  any  reduction  of  prices  will  be 
looked  upon  as  a  proof  of  past  exploitation,  and  that  it  may  be  impossible 
or  difficult  to  increase  price  again.  The  policy  of  letting  sleeping  dogs  lie, 
therefore,  frequently  recommends  itself.  This  applies  to  any  monopoloid 
situation,  but  in  the  case  of  a  cartel  there  are,  fourth,  the  difficulties  of 
taking  action  in  the  face  of  divergent  interests  of  members.  In  Germany, 
for  instance,  it  was  often  the  case  that  the  biggest  and  most  efficient 
firms  opposed  increases  and  sponsored  decreases  of  prices  and  that  the 
smaller  and  less  up-to-date  firms  clamored  for  the  former  and  fought  the 
latter,  bitterly  accusing  the  big  men  of  being  "apostles  of  abstemious- 
ness" (the  writer  is  trying  to  convey  the  meaning  of  Maessiglceitsapostel). 
There  were  many  reasons  for  this.  One  of  them,  however,  was  that  the 
big  firms  did  not  relish  the  idea  of  having  to  back  a  policy  which  pre- 

1  The  above  must  not,  however,  be  interpreted  as  a  defense  of  any  particular  policy. 
All  it  means  is  that  there  are  cases  in  which  measures  of  a  monopolistic  flavor  cannot  be 
condemned  on  traditional  antimonopolistic  grounds.  See,  however,  Professor  Karl 
Pribram,  Controlled  Competition  and  the  Organization  of  American  Industry,  Quarterly 
Journal  of  Economics  for  May  1985. 


540  BUSINESS  CYCLES 

vented  them  from  using  their  powers  to  the  full  and  tended  to  preserve 
economic  lives  the  necessity  of  which  they  were  unable  to  see.  Under- 
standably, this  often  spelled  deadlock,1  with  the  result  that,  no  decision 
being  arrived  at,  prices  did  not  change  at  all  or  changed  with  a  perfectly 
erratic  lag. 

On  the  whole,  readiness  to  combine  is  more  marked  in  Kondratieff 
downgrades  than  in  Kondratieff  prosperities.  (For  examples  see  Chap. 
VII,  especially  sec.  D).  But  there  are  many  exceptions  to  this  general- 
ization. It  seems  that  the  structure  and  conditions  peculiar  to  the 
individual  industry  count  for  more  than  does  the  general  business  situa- 
tion. Nor  is  this  astonishing.  When  things  look  bright,  one  of  the 
strongest  motives  may  be  lacking  for  firms  to  take  shelter,  but  others  are 
gaining  in  importance  and  there  is  less  motive  for  breaking  away  in 
despair — vice  versa  in  depression.  Now  since  an  organization  that 
really  constitutes  anything  at  all  approaching  monopoly,  will,  as  we  have 
seen,  powerfully  influence  the  price-quantity  pairs  of  the  industry,  its 
foundation  or  breakdown  will  dominate  their  statistical  picture  and  may 
even  blot  out  the  influence  of  cyclical  fluctuations  altogether.  Examples 
abound  and  need  not  here  detain  us.  There  is  nothing  in  this  to  modify 
our  theory  of  cycles  or  the  expectations  that  follow  from  it.2 

If  a  cartel  breaks  down,  or  if  a  quasi-equilibrium  among  oligopolists 
is  disturbed,  it  does  not  follow  that  we  may  now  substitute  the  competi- 
tive schema  to  the  one  that  applied  before.  On  the  contrary,  there  will 
be  what  we  call  a  Disorganized  Market.  We  shall  observe  those  moves 
and  countermoves,  which  have  little  relation  to  costs  and  may  have 
almost  as  little  to  the  cyclical  situation,  particularly  if,  during  the  rule  of 
a  cartel,  excess  capacity  has  developed:  bellum  omnium  contra  omnes. 

1  So  did  other  things  which  are  very  real,  though  not  always  easy  to  establish  beyond 
doubt.     There  is  a  committee-room  etiquette,  departure  from  which,  even  if  unintentional, 
may  be  felt  to  be  an  outrage,  driving  the  offended  person  into  restive  ways  which  we  should 
unhesitatingly  call  feminine  if  we  observed  them  at  a  tea  party.     There  are  armchairs  and 
ordinary  chairs.     A  history  of  the  secrets  of  big  business  is  much  more  likely  to  bring  out 
the  importance  of  this  class  of  fact  than  to  reveal  dark  deeds  and  profound  plans. 

2  Responsibility  for  a  large  fixed  investment  may — though  mostly  it  does  not — account 
for  a  wish  to  reduce  risks  by  combination;  necessity  of  committing  oneself  to  large  fixed 
investment  may  account  for  the  existence  of  a  monopoloid  situation.     In  either  case,  there 
is  a  connection,  at  one  remove,  between  size  of  overhead  and  some  motives  for  a  policy  of 
price  stabilization.     But  it  is  perhaps  not  superfluous  to  say  that  there  is  no  other.     The 
mere  fact  of  relative  or  absolute  largeness  of  overhead  would  not,  in  itself,  make  prices  more 
stable  than  they  otherwise  would  be;  it  would,  if  anything,  tend  to  make  them  less  so.     The 
contrary  opinion  would  be  nothing  but  an  error,  if  it  did  not  acquire  some  importance  from 
the  fact  that  business  practice  often  shares  it.     But  observe  that  the  presence  of  the  error 
in  the  practitioner  is  not  proved  by  the  facts,  either  that  he  uses  the  fixed  cost  argument  in 
justification  of  his  refusal  to  reduce  prices  in  depression  to  marginal  costs  or  that  actually 
he  does  not  do  so. 


PRICES  AND  QUANTITIES  OF  INDIVIDUAL  COMMODITIES    541 

The  American  paper  industry  and,  to  a  certain  extent,  the  American 
shoe  industry  in  the  twenties  of  this  century,  or,  at  various  junctures, 
the  German  cement  industry,  may  be  cited  as  instances.  But  such  a 
situation  will  not  always  be  faithfully  reflected  in  the  behavior  of  price- 
quantity  pairs;  for  there  will  be  rebates,  credits,  and  concessions  in  other 
conditions  of  supply.  Changes  in  quality  will  replace  changes  in  prices 
and  make  it  difficult  to  follow  up  changes  in  quantities.  In  fact,  a 
situation  like  this  can  persist  for  some  time  without  affecting  the  quota- 
tions from  which  statistics  are  derived.  This  is  a  case  of  spurious  or 
statistical  stability. 

This  analysis  also  applies  to  the  case  in  which  it  is  a  new  commodity 
that  supplies  the  upsetting  impulse,  although  the  economic  meaning  of 
the  ensuing  competitive  struggle  is  profoundly  different,  the  position  of 
competitors  being  not  unstable  merely,  but  inherently  untenable.  Only 
two  points  call  for  additional  comment.  First,  for  a  considerable  time 
during  which  the  new  article  is  vigorously  gaining  ground,  its  price  as 
well  as  its  quantity  may  be  very  little  sensitive  to  cyclical  fluctuations. 
Demand  may  go  on  shifting  upward  through  several  consecutive  depres- 
sions of  the  Kitchin,  possibly  even  of  the  Juglar,  and  there  may  be  no 
reason  for  the  innovating  firms  to  change  their  prices.  There  are  many 
instances  of  this  in  such  fields  as  the  motorcar  (carrying  with  it  gasoline, 
some  rubber  goods,  nonshatterable  glass,  and  so  on),  electrical  apparatus, 
harvesting  machinery,  rayon,  alloy  steels,  motion  pictures,  and  other 
industries.  Needless  to  say,  such  behavior  is  for  us  anything  but  con- 
trary to  expectation,  however  much  it  may  deviate  from  average  behavior. 
Here,  again,  stability  of  price  is  in  a  sense  spurious.  For  prices  are  stable 
not  primarily  because  they  are  kept  so,  but  because  opposing  forces, 
acting  upon  price  and  quantity  make  them  so:  another  instance  to  show 
that  prices  which  do  not  change  for  a  time  are  not,  ipso  facto,  rigid  in 
our  sense.1 

The  second  point  refers  to  those  situations  in  the  cyclical  process  of 
evolution  in  which  entrepreneurs  temporarily  enjoy  what,  on  a  previous 
occasion,  has  been  called  an  acceptable  approximation  to  straight 
monopoly  (Chap.  II,  sec.  F,  III).  In  such  situations,  several  courses  are 
open  to  them.  For  instance,  they  may  decide  to  make  hay  while  the 
sun  shines,  to  use  this  hay  to  write  off  their  plants  to  $1  as  quickly  as 
possible,  and  accept  defeat  without  struggle  when  their  demand  curve 
crumbles.  It  should  be  observed  that  in  such  truly  monopolistic  cases 
it  would,  in  general,  precisely  not  be  to  the  interest  of  a  seller  to  keep  his 
prices  stable  through  changing  situations.  But  as  a  rule  an  entrepreneur 
has  no  such  chance  and  no  such  intention.  He  must  build  up  demand 

1  See,  however,  V.  A.  Mund,  Prices  under  Competition  and  Monopoly,  Quarterly  Journal 
of  Economics  for  February  1984. 


542  BUSINESS  CYCLES 

and  then  defend  the  ground  conquered  against  the  attacks  of  competitors 
for  whom  he  fatally  paves  the  way.  Hence,  he  is  neither  before  nor  after 
success  in  a  position  to  behave  according  to  the  schema  of  the  classical 
theory  of  monopoly.  In  those  cases  in  particular,  which  have  steadily 
grown  in  importance  ever  since  Huntsman  produced  steel  in  the  dark  of 
the  night,  by  the  work  of  his  own  hands,  in  which  the  entrepreneurial 
achievement  consists  in  or  presupposes  the  creation  of  facilities  for  mass 
production,  demand  can  be  built  up  and  the  ground  conquered  be 
defended  only  by  selling  cheaply  from  the  start  and  never  raising  prices 
afterward.  Visualizing  the  price  that  in  the  average  of  good  and  bad 
years  will  attain  both  ends,  and  being  able  to  produce  at  costs  which 
that  price  will  more  than  cover  is,  in  those  cases,  the  main  requisite  of 
success.  Theoretically,  such  a  price  need  not  be  inflexible  downward, 
but  it  is  easy  to  see  that  under  the  circumstances  any  reduction,  unless 
based  on  quite  permanent  conditions,  must  be  a  dangerous  step  to  take. 
Hence  we  find,  mainly  in  the  field  of  highly  finished  consumers'  goods  but 
also  in  the  field  of  tools  and  machinery,  so  frequently  prices  that  according 
to  all  ordinary  standards  are  extremely  rigid — the  price  becoming  part  of 
the  individuality  of  the  commodity  and  the  firm  being  practically  com- 
mitted to  it.  But  the  point  to  be  made  is  that  this  kind  of  rigidity 
differs  from,  and  does  not  entail  the  same  cyclical  consequences  as, 
rigidity  in  the  usual  sense.  The  latter  is  defined  by,  and  its  effects  turn 
on,  deviation  from  the  price  that  would  at  any  moment  be  ideally — never 
mind  now  whether  in  a  monopoly  or  competitive  sense — adapted  to  the 
general  situation. 

A  price  of  the  kind  envisaged  will  not  conform  to  that  standard,  but 
it  will,  since  it  has  been  fixed  with  a  view  to  an  expected  sequence  of 
situations,  come  in  case  of  success — which  of  course  is  a  matter  of  a  lucky 
throw — much  nearer  to  it  than  a  price  fixed,  say,  by  compromise  such  as  is 
characteristic  of  a  cartel  or  by  some  rule  dictated  by  a  public  service  com- 
mission. It  is  also  not  true  that  in  these  cases,  as  a  well-known  slogan 
has  it,  "quantity  adapts  itself  to  price  instead  of  price  to  quantity," 
because  considerations  about  future  quantities  of  output  precisely  hold 
first  place  in  determining  the  decision.  Moreover,  it  must  not  be  for- 
gotten that  any  new  article,  however  inflexible  its  price  may  be,  is  still 
a  tool  of  flexibility  in  the  system  as  a  whole,  by  virtue  of  its  impact  on 
the  preexisting  price-quantity  structures  in  its  field.  Finally,  if  adjust- 
ment becomes  necessary  after  all,  there  will  be  motive  to  effect  it  through 
change  in  quality  or,  if  that  be  impossible  or  undesirable,  by  offering 
another  type  or  many  other  types  of  product  at  different  prices.  Statis- 
tics will  then,  if  the  old  type  is  not  discarded,  record  absolutely  rigid 
prices  for  both  the  old  and  the  new  type,  while  to  all  intents  and  purposes 
there  is — an  inhibited,  no  doubt,  and  discontinuous — flexibility  and  the 


PRICES  AND  QUANTITIES  OF  INDIVIDUAL  COMMODITIES    543 

usual  inferences  from  price  rigidity  are  largely  unwarranted,  though 
of  course  not  wholly  so.  This  is  particularly  clear  if  several  similar 
articles  are  offered  from  the  start — "a  car  for  every  purse,"  for  instance — 
when  no  actual  change  of  price  is  necessary  at  all  in  order  to  produce 
practically  all  the  effects  of  prompt  variability. 

Thus,  analysis  of  the  nature  and  sources  of  the  various  kinds  of  price 
rigidity  we  observe  and  of  that  monopolistic  or  oligopolistic  strategy 
which,  intentionally  or  unintentionally,  rationally  or  irrationally,  is 
responsible  for  some  of  them,  hardly  lends  support  to  the  ideas  many 
students  entertain  about  their  importance  or,  as  some  would  say,  growing 
importance  for  the  cyclical  mechanism,  particularly,  their  dislocating 
effects  on  the  rest  of  the  system  in  depression.  There  is  less  genuine 
rigidity,  and  what  there  is  of  it  is  less  dislocating,  than  is  widely  assumed. 
We  should  not,  however,  run  into  the  opposite  error.  The  very  fact  that 
every  concern  which  has  any  opportunity  to  do  so  strives  to  secure  a  posi- 
tion that  looks  as  monopolistic  as  possible — by  advertising,  differentiating 
its  product,  trying  to  control  competitors  (electric  power  and  light  com- 
panies, for  example,  tried  to  get  control  of  gas  companies;  American  rail- 
roads tried  to  buy  up  trolley  lines),  warding  off  competition  by  preventive 
attack — is  in  itself  sufficient  to  prove  that  such  positions  are  not  valueless. 
But  we  have  seen  in  our  historical  sketch  that  their  value  consists, 
rather  than  in  any  power  to  follow  a  long-run  policy  of  restriction  of 
output,  in  the  facilities  they  afford  for  steering  safely  through  difficult 
situations  and  for  undisturbed  planning.  And  the  difference  the  existence 
of  such  positions  makes  is  not  so  much  a  difference  in  ultimate  results 
as  a  difference  in  the  way  by  which  they  are  reached.  The  latter  is,  to  be 
sure,  quite  sufficient  to  upset  our  expectations  as  to  the  behavior  of  price- 
quantity  pairs  in  the  short  run,  i.e.,  practically  in  the  Kitchin.  But  it  is 
not  sufficient  to  upset  the  working  of  our  process. 


CHAPTER  XI 


Expenditure,  Wages,  Customers' 
Balances 


A.  Some  Propositions  about  Money. — Propositions  about  money 
have  been  introduced  already,  and  others  will  be  when  necessary.  Most 
of  them  are  to  be  found  in  Chaps.  Ill,  VIII,  and  XIII.  Still  others 
will  be  assembled  in  this  section.  Since  it  is  not  possible  to  deal  in 
this  book  with  the  general  theory  of  money  to  which  these  propositions 
pertain,  they  must  here  appear  in  an  incomplete  and  unsatisfactory  form 
and  detached  from  the  background  on  which  alone  they  could  acquire 
their  full  meaning.1 

1.  Money  may  be,  and  in  practice  mostly  is — at  least  historically — 
linked  to  some  commodity.  But  it  never  is  a  commodity  and  never 
satisfies  wants  in  the  sense  in  which  commodities  do.  If  we  nevertheless 
attribute  utility  to  it,  this  utility  is  derived  from  that  of  the  commodities 
we  actually  buy,  or  could  buy,  with  it  and  hence  presupposes  given  prices, 
or  ratios  of  exchange  between  money  and  commodities.  Any  attempt, 
therefore,  to  deduce  these  prices  from  marginal  utilities  of  money  and 
marginal  utilities  of  commodities,  in  the  way  in  which  we  may  deduce 
exchange  ratios  between  commodities  from  their  marginal  utilities, 
involves  circular  reasoning.  This  is  so  even  if  the  monetary  unit  consists, 
for  instance,  of  a  given  quantity  of  metal  that  can  be  freely  coined  and 
melted  without  cost  or  loss  of  interest.  For  although  in  this  case  the 
exchange  value  of  that  quantity  of  metal  in  its  monetary  use  can  never 
depart  from  its  exchange  value  in  its  industrial  uses,  the  former  cannot 
fundamentally  be  explained  by  the  latter :  the  closing  of  the  mints  to  the 
public  would  suffice  to  bring  out  the  fact  that  money  has  an  exchange 
value  of  its  own,  which,  therefore,  as  long  as  coinage  is  unrestricted, 
"determines"  the  exchange  value  of  the  money  commodity  in  its  indus- 
trial uses  quite  as  much  as  it  is  "determined"  by  it.  Unfortunately  we 
cannot  enter  into  the  discussion  of  the  various  devices  by  which  theorists 
have  labored  to  avoid  the  implications  of  this.  It  is  hoped,  however,  that 
our  argument  is  clear  as  it  stands.  If  so,  we  may  draw  the  conclusion 
that  any  kind  of  linking  of  the  monetary  unit  to  the  unit  of  a  commodity, 

1  The  writer  hopes  to  provide  that  background  and  to  develop  the  theoretical  structure 
of  which  these  propositions  are  fragments,  in  his  treatise  on  money. 

544 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES      545 

whatever  its  practical  merits  in  guaranteeing  the  value  of  money  may  be, 
is  logically  nonessential  and  subjects  the  functioning  of  the  monetary  sys- 
tem to  an  additional  condition  which  is  extraneous  to  the  meaning  of 
money. 

2.  The  fact  that  money  is  not  a  commodity  explains  what  otherwise 
would  be  inexplicable,  namely,  that  claims  or  titles  to  money  (however 
defined)  may  serve  the  same  purposes  as  money  itself.     This  is  the  funda- 
mental explanation  of  the  possibility  of  "credit  creation,"  as  well  as  the 
reason  why   it  is  so  easy  to  create  "near  money"1  and  so  difficult  to 
prevent  the  creation  of  it.     No  such  thing  can  occur  in  the  case  of  a 
commodity. 

3.  The  same  fact  underlies  the  phenomenon  of  "velocity  of  money," 
which,  similarly,  has  no  analogy  in  the  world  of  goods  and  services. 
This  phenomenon  is  not  correctly  described  by  calling  money  a  durable 
good  which  can  serve  many  times.     Money  serves  merely  as  a  counter, 
which,  within  technical  limits,  can  turn  up  any  number  of  times  during 
the  game.     These  limits  are  essential,  however.     The   "periods,"  i.e., 
the  spans  of  time  which  it  takes  the  monetary  units  to  complete  their 
circuit,  are  the  fundamental  facts  about  monetary  circulation. 

We  have  to  distinguish  three  concepts  of  velocity:  (a)  the  sum  total 
of  all  transactions  in  terms  of  money  divided  by  checking  balances  plus 
money  outside  of  banks;  (6)  consumers'  plus  producers'  expenditure  by 
balances  plus  money  in  circulation;  (c)  consumers'  expenditure  by 
balances  plus  money  in  circulation.  The  last,  the  so-called  income 
velocity — we  may  call  it  net  income  velocity  to  distinguish  it  from  gross 
income  velocity  equal  to  total  income  by  balances  plus  money  in  circula- 
tion— is  the  most  relevant  figure  of  the  three;  the  first  is,  in  itself,  mean- 
ingless, though  it  must  sometimes  serve  as  the  only  indicator  we  have 
of  the  other  two. 

More  important,  however,  is  another  distinction,  which  may  be 
applied  to  any  of  the  three  velocity  figures  just  mentioned.  In  a  station- 
ary state,  velocity  would  be  largely  determined  by  the  institutional 
arrangement  of  payments  within  the  period  of  account.  Suppose  the 
economic  process  to  be  organized  in  such  a  way  that  all  firms  buy  produc- 
tive services  from  households,  say,  on  each  Saturday,  and  all  households 
buy  consumers'  goods  from  the  same  firms  and  for  the  same  amount  of 
money,  on  each  subsequent  Monday.  This  being  all  that  happens, 
income  velocity  per  year  is  52.  We  may  go  a  step  further,  however, 
and  assume  that  incomes  are  paid  out  and  spent  concurrently  all  through 

1  The  writer  supposes  that  that  term  has  been  coined  by  way  of  analogy  with  "near 
beer."  If  so»  it  does  not  express  the  whole  truth.  Near  beer  did  not  serve  so  well  as  beer 
and  it  had  to  be  produced.  It  is,  also,  not  quite  correct  to  say  that  "credit"  serves  as  a 
"substitute  for  money." 


546  BUSINESS  CYCLES 

the  week  in  small  installments,  and  that  the  sequence  of  everyone's  indi- 
vidual payments  and  receipts  is  random.  This  case  will,  of  course,  give 
a  different  velocity  and,  moreover,  display  an  element  absent  in  the  first 
case.  Households  as  well  as  firms  will  now  hold  an  element  of  cash 
specifically  intended  to  provide  for  the  occurrence  of  unfavorable 
sequences.  Subject  to  this  new  proviso,  however,  there  is  still  a  deter- 
mined velocity  measuring  the  way  through  economic  space  of  every  unit 
of  money  spent — and  all  units  that  are  spent  are  assumed  to  be  promptly 
spent,  for  there  is,  in  this  setup,  no  object  in  withholding  them.  We 
shall  use  the  word  efficiency  for  this  kind  of  velocity,  or  confine  the  term 
velocity  to  it.  For  it  the  classical  assumption  of  constancy  or  slow  and 
independent  variation  is  approximately  true,  not  only  in  the  stationary 
state,  but  in  all  cases,  that  of  extreme  inflation  alone  excepted. 

As  soon,  however,  as  we  leave  the  precincts  of  the  stationary  process, 
in  order  to  deal  with  changing  business  situations,  we  meet  besides  effi- 
ciency a  phenomenon  which  is  entirely  different  from  it,  although  it 
influences  velocity  figures  similarly.  We  find  that  people  sometimes  do 
withhold  money  they  intended,  and  still  intend,  to  spend;  and  that, 
whereas  before  spending  was  a  matter  of  course,  the  question  whether  or 
not  to  spend  at  any  moment  becomes  a  question  of  policy  for  everyone, 
obviously  important  for  the  picture  the  monetary  process  will  present. 
While  efficiency  refers  to  the  velocity  of  any  unit  that  is  actually  sent  over 
its  path,  we  now  find  another  component  of  the  velocity  figure  which 
refers  to  the  proportion  of  existing  units  so  sent.  We  shall  call  it  Rate 
of  Spending,  or  simply  Spending  (see  Chap.  Ill,  sec.  A),  and  shall  put 
its  equilibrium  (stationary)  value  as  equal  to  unity.  Obviously,  this  is 
the  cyclical  variable  within  total  velocity. 

4.  Since  certain  claims  to  "money"  serve,  within  wide  limits,  the 
same  purposes  as  legal  tender  itself,  it  is  not  only  necessary  to  include  the 
existing  amount  of  such  claims  in  the  total  quantity  of  money  (typical 
cases:  bank  notes,  deposits  subject  to  check),  but  also  evident  that 
the  very  concept  of  quantity  of  money  becomes  doubtful.  It  is,  in  fact, 
impossible  to  speak  of  the  quantity  of  "  money  "  in  the  sense  in  which  we  speak 
of  the  quantity  of  a  commodity.1  We  therefore  stand  to  lose  what,  as  we 
have  seen  in  the  chapters  on  price  level  and  prices,  is  nevertheless  a 
necessary  element  in  the  determination  of  values.  What  is  to  fill  this 
breach  in  the  theory  of  money  is  a  problem  which  cannot  be  attacked  here. 

In  any  case,  however,  we  cannot  here  consider  quantity  of  "  existing  " 
or  "circulating"  or  "available"  money  as  an  independent  variable,' 
because,  although  it  varies  in  function  of  some  elements  that  may,  in  the 

1  Another  consequence  is  that  the  distinction  between  velocity  and  quantity  becomes 
blurred.  K.  Wicksell,  for  instance,  treated  the  issue  of  bank  notes  as  a  means  of  increasing 
not  the  quantity  but  the  velocity  of  money  (the  banks'  reserves). 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES      547 

sense  of  the  old  quantity  theory,  be  looked  upon  as  data,  it  also  varies  in 
response  to  other  variables  of  our  process,  entrepreneurial  activity  in 
particular.  It  has  been  pointed  out  before  that  many  modern  authors — 
most  of  them  adherents  of  the  investment  theory  of  banking  (Chap.  Ill) 
— try  to  meet  this  situation  by  what  may  be  interpreted  as  a  revival  of 
the  quantity  theory.  They  figure  out  what  the  technical  maximum  of 
credit  creation  is  within  a  given  institutional  frame — legal  or  traditional 
rules  about  reserve  proportions  and  the  like — attribute  to  banks  a 
tendency  to  maintain  that  maximum  of  customers*  balances  and  thus 
construct  a  quantity  of  money  which  is  held  to  act  as  such  on  the  economic 
process  (and  to  have  a  causal  significance)  in  much  the  same  way  as  the 
quantity  of  money  in  the  quantity  theory  sense  was  by  older  authors  held 
to  act.  In  Chap.  XIII  we  shall  see  how  far  from  satisfactory  any  such 
mechanistic  theory  of  banking  is.  Here  it  is  sufficient  to  state  that,  even 
disregarding  the  fact  that  that  institutional  frame  is  not  independent  of 
the  evolutionary  process,  the  maximum  alluded  to  does  not  constitute 
the  supply  but  at  best  the  limit  of  the  supply  of  balances,  and  that  the 
amount  banks  are,  in  a  given  situation,  willing  to  supply  cannot  be 
explained  by  considerations  at  all  analogous  to  those  that  are  applicable 
in  the  case  of  the  supply  of  a  commodity. 

5.  Money  not  being  a  commodity,  the  traditional  apparatus  of  supply 
and  demand  cannot  be  applied  to  the  solution  of  the  problem  of  money 
prices  of  commodities  and  of  the  price  levels  (see  Chap.  VIII).  An 
exchange  of  money  for  commodities  is  not  the  same  phenomenon  as  an 
exchange  of  one  commodity  against  another.  This  is  still  clearer  on  the 
demand  side  than  it  is  on  the  supply  side.  We  may,  with  some  qualifica- 
tions, speak  of  a  demand  for  money  in  the  money  market.  But  there 
is  no  sense  in  speaking  of  a  demand  for  money  displayed  by  sellers  in  a 
commodity  market.  Demand  for  money  carries,  however,  still  another 
meaning:  it  may  mean  the  wish  to  hold  stocks  of  money  or  balances. 
This  Walrasian  idea  of  an  encaisse  dSsirSe,  which  reappears  in  Marshallian 
analysis  and  has  recently,  it  seems,  entered  upon  a  new  lease  of  life,  is  one 
of  the  least  valuable  elements  in  the  great  Frenchman's  mighty  structure. 
It  is  harmless  only  in  the  analysis  of  stationary  states,  although  even  there 
it  implies  a  misrepresentation  of  facts.  If  people  get  their  "incomes" 
each  Saturday  and  spend  them  on  consumers'  goods  each  succeeding 
Monday — transactions  between  firms  being  excluded — then  the  money 
will  lie  about  in  the  vaults  of  firms  from  Monday  to  Saturday,  not 
because  there  is  any  demand  for  cash  holdings,  but  because  the  institu- 
tional arrangement  so  wills  it.  But  the  idea  becomes  misleading  if  we 
leave  the  stationary  case.  If  we  see  someone  displaying  a  wish  for  bread, 
this  is  a  clear-cut  fact  carrying  its  explanation  in  itself  and  fit  to  be  used 
in  order  to  deduce  the  explanation  of  other  facts.  But  if  someone  dis- 


548  BUSINESS  CYCLES 

plays  a  wish  to  hold  cash,  this  in  itself  means  nothing  at  all.  All  the 
value  of  the  observation  lies  in  the  circumstances  that  induce  that  wish — 
all  the  theory  of  the  fact  and  its  consequences  turns  upon  those  circum- 
stances— even  if  there  is  such  a  wish.  But  generally  there  is  none.  A 
man  may,  for  example,  hold  a  supernormal  amount  of  cash,  not  because 
this  is  any  good  to  him,  but  simply  because  his  and  other  peoples'  actions 
happen  to  produce  that  result,  which  in  itself  is  not  one  of  the  objects  he 
wishes  to  attain  by  those  actions;  it  may  even  be  a  disagreeable  by-prod- 
uct of  them.  All  explanations  which  start  with  the  famous  adage:  "If 
people  choose  to  hold  •  •  •  "  are  ipso  facto  condemned. 

6.  A  general  remark  may  conveniently  be  inserted  here.  One  of  the 
first  tasks  that  confronted  scientific  economics  in  the  early  stages  of  its 
career  was  to  fight  certain  popular  views  about  money  which,  however 
understandable  or  even  defensible  they  may  now  seem  to  the  historian, 
were  in  fact  largely  erroneous.  During  their  campaign  against  "bullion- 
ist"  and  "mercantilist"  exaggerations  of  the  importance  of  the  role  of 
money,  economists  were  naturally  driven  to  construct  a  body  of  doctrine 
in  real  terms  alone.  They  tried  to  draw  away  the  "veil"  of  money  in 
order  to  describe  the  process  of  the  production  and  consumption  of 
wealth.  This  effort,  although  highly  meritorious  at  the  time,  was  bound 
to  fail  in  the  end.  Economic  action  cannot,  at  least  in  capitalist  society, 
be  explained  without  taking  account  of  money,  and  practically  all 
economic  propositions  are  relative  to  the  modus  operandi  of  a  given 
monetary  system.  In  this  sense  any  theory  of,  say,  wages  or  unemploy- 
ment or  foreign  trade  or  monopoly  must  be  a  "monetary"  theory,  even 
if  the  phenomenon  under  study  can  be  defined  in  nonmonetary  terms. 
This  is  increasingly  being  recognized,  and  the  fact  that  it  is  must  be  listed 
among  the  major  improvements  our  analytic  apparatus  has  undergone 
during  the  last  20  years  or  so.  But  we  seem  unable  to  draw  away  from 
an  old  error  without  running  into  the  opposite  one,  which  in  this  case  is 
older  still.  That  economic  analysis  cannot — in  the  sense  in  which  for 
instance  Boehm-Bawerk  thought  it  could — abstract  from  money  is  a 
truth  which  is  useful  only  if  supplemented  by  the  other  truth  that 
monetary  processes  never  carry  their  explanation  in  themselves  and  can- 
not be  analyzed  in  monetary  terms  alone.  And  recognition  of  the  fact 
that,  in  its  fight  against  mercantilist  views  about  the  causal  role  of  money 
in  economic  life,  classical  doctrine  went  much  too  far  must  be  supple- 
mented by  the  other  fact  that  in  fighting  them  it  also  performed  a  service 
of  which  we  again  stand  in  need  today. 

B.  System  Expenditure  (Outside  Clearings). — The  changes  in 
monetary  expressions  and  monetary  quantities  that  occur  in  the  course 
of  the  cyclical  process  of  evolution  either  reveal  themselves  in,  or  are 
brought  about  by,  changes  in  the  totals  of  business  and  household 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES      549 

expenditure,  which  are — and  so  far  we  agree  with  the  most  monetary  of 
theories  of  the  cycle — the  most  important  immediate  conductors,  as  well 
as  the  most  obvious  effects,  of  changes  in  the  general  complexion  of  busi- 
ness situations.  In  a  general  theory  of  money  it  would  evidently  be 
appropriate  to  consider  producers'  and  consumers'  expenditure  as 
mutually  interdependent:  producers'  expenditure  expands  and  contracts 
in  function  of  consumers'  expenditure  exactly  as  consumers'  expenditure 
expands  and  contracts  in  function  of  producers'  expenditure,  that  is  to 
say,  in  response  to  the  changes  in  income  paid  by  firms  to  households. 
But  for  the  particular  process  with  which  we  are  concerned  here,  it  is 
more  helpful  to  recognize  that  the  fundamental  impetus  comes  from 
producers'  (entrepreneurs')  expenditure  and  that  households  merely 
react  to  it. 

It  is  necessary  to  use  this  proposition  with  care.  Instances  of  the 
reverse  relation  are  not  lacking,  even  within  our  process:  that  reaction 
of  the  business  world  which  makes  up  what  we  have  called  the  Secondary 
Wave  is  partly  reaction  by  firms  to  increased  consumers'  expenditure. 
But  inasmuch  as  these  phenomena  can  be  traced  back  to  the  impulse 
given  by  entrepreneurs'  expenditure  made  in  the  course  of  carrying  out 
new  combinations  of  factors,  they  may  also  be  considered  as  consequences 
of  that  prime  mover.  Not  all  incomes,  moreover,  are  paid  out  by  firms, 
nor  do  all  that  are,  vary  with  the  sum  total  of  producers'  expenditure. 
Exceptions  must  be  made  for  them.  But,  with  these  qualifications  and 
in  this  sense,  it  is  still  broadly  true  that  household  expenditure  varies  in 
function  of  producers'  expenditure,  although,  of  course,  not  simply  in 
direct  proportion  to  it.  Common  sense  and  common  experience  will 
probably  find  little  fault  with  the  simplifying  generalization  that  pro- 
ducers' expenditure  is,  within  our  process,  the  active  element  of  total 
expenditure  in  the  system  (consumers'  plus  producers'  expenditure)  or, 
to  use  a  term  due  to  Mr.  C.  E.  Thomas,  of  system  expenditure.  We  shall 
come  nearer  to  facts  if  we  assume  this  quantity  to  vary,  not  only  in  func- 
tion of  producers'  expenditure,  but  also  in  function  of  its  rate  of  change, 
which  takes  care  of  the  most  common  type  of  anticipations  as  to  the  imme- 
diate future:  producers'  expenditure  molds  consumers'  expenditure,  not 
only  by  supplying  households  with  money  to  spend,  but  also  by  shaping 
their  willingness  to  spend;  and  this  willingness  largely  depends  on  the 
rate  of  change  of  income  streams  which  prevails  at  the  moment.1  The 

1  Even  the  relation,  per  period  of  account,  between  firms'  expenditure  and  households, 
receipts  is,  owing  to  lags  and  to  payments  between  firms,  not  simple  or  invariant,  let  alone 
one  of  identity.  Still  more  important  is  it  to  observe  that,  although  there  is  no  doubt  about 
the  direction  of  the  influence  of  household's  receipts  on  household's  expenditure  on  con- 
sumers' goods,  there  is  plenty  of  doubt  about  the  form  of  this  function  also.  It  may  vary 
widely  as  between  nations  and  situations.  The  French  peasant  and  small  bourgeois,  as  well 
as — perhaps — the  "Scotchy"  New  Englander  of  old,  may  provide  instances  of  a  limiting 


550  BUSINESS  CYCLES 

question  of  introducing  a  lag — which  would  in  any  case  be  a  short  one — 
need  not  be  touched  at  our  stage  of  approximation. 

The  most  comprehensive  series  of  expenditure  is  the  series  of  total 
debits  to  banking  accounts.  In  the  United  States  this  total  may  be  said 
to  approach  the  sum  of  all  monetary  transactions,  although  an  amount  of 
transactions  still  remains  unrecorded  which  it  is  hardly  possible  to  esti- 
mate accurately  and  which  of  course  varied  greatly  in  the  slow  evolution  of 
banking  habits  during  150  years  before  the  war.  It  should  be  added  that, 
exactly  as  some  monetary  transactions  fail  to  show  up  in  debits,  so  many 

case  in  which  increase  in  receipts  entirely  fails  to  induce  an  increase  in  consumers'  expendi- 
ture. Again,  sudden  increase  in  monetary  receipts,  due,  for  instance,  to  the  incipient  stages 
of  inflation,  may  sometimes  also  fail  to  produce  a  correspondingly  great  effect — that  is  why 
prices  rise  less  than  balances  or  quantity  of  fiat  money  during  those  stages  or  why  balances 
themselves  rise  less  than  in  proportion  to  government  spending,  the  new  access  of  wealth 
being  partly  applied  to  the  repayment  of  debt.  But  opposite  cases  seem  to  the  writer, 
under  ordinary  circumstances,  to  be  much  more  frequent  and  important.  The  working 
hypothesis  which  we  shall  mainly  rely  on  is,  hence,  that  consumers'  expenditure  increases 
more  than  proportionately  whenever  household  receipts  increase  and  decreases  more  than 
proportionately — particularly  if  durable  consumers'  goods  play  a  great  relative  role  in 
consumers'  budgets — whenever  household's  receipts  decrease.  This  comes  practically 
to  saying  that  there  is  a  general  tendency  for  the  average  household  to  outrun  its  "income" 
in  all  phases  of  all  cycles  excepting  depression,  in  which,  however,  forced  dissaving  and 
subsidies  counteract  the  effect  to  some — historically  variable — extent.  This  hypothesis 
rests  on  the  following  observations,  which  however,  it  will  be  observed,  are  all  drawn  from 
"modern  times" — roughly,  from  the  time  since  1900. 

1.  The  writer  thinks  that  he  has  indeed  observed  a  type  of  industrial  family  (and,  in 
some  countries,  of  peasant  family,  as  stated  above),  though  much  more  clearly  in  Austria, 
England,  France,  and  Germany  than  in  this  country,  which  keeps  to  a  very  rigid  style  of 
life,  mainly  expressed  in  its  family  houses  and  the  way  they  are  run,  and  varies  its  expendi- 
ture but  little  as  between  prosperity  and  depression.     The  vast  majority  of  all  classes 
behaves  differently.     One  stratum  (witness  the  strongly  cyclical  character  of  the  jewelry, 
theater,  and  hotel  trades)  obviously  spends  the  speculative  and  other  temporary  gains — 
the  former  are  not  income  in  our  sense — of  prosperity  on  consumers'  goods.     The  wage- 
earning  stratum  seems  to  the  writer  to  display  the  same  bent  and  to  try  in  times  of  rising 
wages  to  spend  "all  it  can."     This  is  true  even  for  noncyclical  increases  of  income,  such 
as  periodically  occur  in  the  salaries  of  state  employees;  in  Germany,  at  least,  they  seemed  to 
spend  more  than  the  increase  each  time  their  salaries  were  augmented. 

2.  All  those  impressions  are  supported  by  the  growth  of  the  practice  of  buying  on 
installments,  although  this  practice  would  not,  in  itself,  suffice  to  establish  them. 

8.  It  is  a  fact  that  the  general  net  indebtedness  of  households  increases  in  times  of 
rising  incomes,  though  this  fact  can  be  fully  proved  for  postwar  times  only. 

4.  The  presence  of  a  tendency  in  the  vast  majority  of  people  to  spend  on  consumers' 
goods  whatever  they  can,  excepting  depression,  is  suggested — though  again  not  strictly 
proved — by  the  monthly  and  weekly  rhythm  in  retail  trade,     (cf.  Gruenbaum,  Umsatz- 
schwankungen  des  Einzelhandels,  Sonderheft  10  of  the  Berlin  Institute's  Vierteljahrshefte, 
1928.) 

5.  The  contrary  impression,  which  so  many  economists  seem  to  have,  can  be  traced 
to  the  one  item  to  which  the  above  clearly  does  not  apply,  viz.,  profits  in  our  sense. 

We  shall  return  to  the  subject  presently,  and  again  in  Chaps.  XII  and  XIV. 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES      551 

commodity  transactions  fail  to  give  rise  to  monetary  transactions — 
especially  the  "transactions"  that  consist  in  producers',  particularly 
farmers',  "selling"  part  of  their  own  produce  to  themselves.  This  debit 
series  is,  however,  not  available  for  the  prewar  time.  Instead,  we  have 
the  series  of  bank  clearings,  which  must  be  expected  to  differ  from  the 
true  figure  of  debits  that  it  is  supposed  to  indicate,  and  even  this  we  have 
only  for  this  country  and  England.  But  the  American  figure  has  been 
shown  to  display  so  close  a  relation  to  the  figure  of  debits  since  the  war, 
that  we  may  with  some  confidence  use  it.  In  countries  and  for  times  in 
which  banking  is  highly  concentrated  the  situation  would,  of  course,  be 
less  satisfactory. 

Apart  from  the  changing  relative  importance  of  the  amount  recorded 
and  various  difficulties  of  a  technical  nature  into  which  we  cannot  enter 
here,  it  has  to  be  recognized  that  the  total  is,  as  such,  a  composite  of  very 
little  significance.  Debits  arising  out  of  transactions  between  banks  have 
a  special  character  and  should  be  excluded,  as  should — for  some  purposes, 
though  not  for  others — debits  to  public  accounts.  Moreover,  the  total 
records  every  charity  and  every  tax  payment,  as  well  as  disbursements 
from  incomes  which  are  not  paid  out  by  firms.  However,  some  of  these 
elements  are  not  important  enough  to  distort  the  picture;  others  are  as  a 
matter  of  fact  excluded.  What  seriously  impairs  the  value  of  the  clear- 
ings series  is  the  impossibility  of  separating  stock  exchange  and  real 
estate  transactions  from  the  rest.  All  we  can  do  about  this  is  to  adopt 
the  familiar  distinction  between  New  York  and  Outside  Clearings, 
although  the  former  contain  all  the  transactions  of  the  world's  greatest 
industrial  and  commercial  center  and  the  latter,  a  very  considerable 
element  of  stock  exchange  business  and  of  speculative  transactions  in 
general.  Similarly  we  must,  in  the  English  case,  pin  our  faith  to  the 
distinction  between  town  clearings  at  the  London  clearinghouse  and 
provincial  clearings  plus  country  clearings  at  the  London  clearinghouse. 

Bearing  in  mind,  then,  that  they  really  measure  something  different 
and,  in  particular,  that  they  record  the  value  of  every  element  of  every 
commodity  as  often  as  it  changes  hands,  against  payment  by  check, 
in  the  course  of  production  and  trade,  we  must  resign  ourselves  to  using 
Outside  Clearings1  (Mr.  Frickey's  series  mainly)  to  indicate  variations 

1  Series  are  available  monthly  for  a  varying  number  of  cities,  which  increased  to  159, 
from  the  Financial  Review,  the  Public,  the  Commercial  and  Financial  Chronicle,  and  other 
sources.  Mr.  Snyder's  compilation  goes  back  to  1875  (Journal  of  the  American  Statistical 
Association,  1924).  He  corrects  by  his  general  price  level.  Mr.  Frickey's  monthly  series 
for  seven  selected  cities  (Baltimore,  Chicago,  Cincinnati,  Cleveland,  Philadelphia,  Pitts- 
burgh, San  Francisco)  is  the  only  one  to  take  account  of  the  fundamental  difficulties  of 
this  material,  consisting  in  the  effect  of  new  clearinghouses  emerging  from  time  to  time, 
before  1903  (see  E.  Frickey's  paper  in  Review  of  Economic  Statistics  for  October  1935). 
Some  experimentation  has  quickly  shown  it  to  be,  because  of  this  and  other  virtues,  the 


552 


BUSINESS  CYCLES 


"NEfDEPOSIT: 
INVESTMENTS 


MINUS 
OUTSIDE  N.Y.C, 


f,''\'C(  EARINGS 


in  system  expenditure.  Some  idea  of  the  extent  of  the  risk  we  run  when 
reasoning  as  if  outside  clearings  represented  the  dollar  volume  of  physical 
production  may,  however,  be  derived  from  inspecting  Chart  XXIII. 
Whatever  confidence  we  may  place  in  the  evidence  it  presents,  it  certainly 
goes  some  way  toward  allaying  extreme  apprehensions.  The  covariation 
between  Outside  Clearings  and  Physical  Output  of  Industry  times  the 

Price  Level  is,  given  the  statistical  in- 
dependence of  the  respective  materials, 
in  itself  an  interesting  result.  Possi- 
bilities of  very  simple  "theories"  seem 
to  open  up  at  this  point. 

System  expenditure  makes  a  natural 
and  systematic  series  of  the  "velocity" 
(or  rate-per-time)  type,  which  is  ob- 
viously cyclical.  It  is  a  primary  but 
consequential  phenomenon.  There  is 
no  need  to  stress  that  it  is  primary :  in- 
crease in  system  expenditure  is  not  only 
one  of  the  most  obvious  elements  in 
the  picture  of  any  prosperity,  but  also 
the  most  important  factor  in  produc- 
ing the  symptoms  that  we  associate 
with  prosperities.  All  the  more  neces- 
sary is  it  to  emphasize  that  it  is  neverthe- 
less consequential:  no  antecedent  or 
initiative  increase  in  system  expendi- 
ture— induced  for  instance  by  some 
chance  event  or  by  the  initiative  of 

banks  or  governments — is  required  in  order  to  start  the  extrepreneurial 
activity  that  propels  the  system  away  from  any  preexisting  neighborhood 
of  equilibrium.  This  activity  directly  implies  and  indirectly  induces  all 
that  additional  spending  we  observe  in  prosperities,  but  the  innovations 
themselves  are  independent  of  it  in  the  sense  that  they  are  profitable  at 
the  system  expenditure  in  the  preexisting  neighborhood.1  If  they  are 


J_ 


/MANUFACTURING 
/AND  MINING  X  PRICES 


_L 


1890        95  1900          05         1910     13 

CHART    XXIII.— United    States    (see 

Appendix,  p.  1059). 


best  one  to  use,  and,  with  one  exception,  it  has  been  primarily  used  by  the  writer  in  his 
work  on  system  expenditure. 

1  See  the  analogous  statement  about  the  price  level  (Chaps.  IV  and  VIII).  There  is, 
thus,  so  far  neither  necessity  nor  room  for  the  assumption  that  either  "spending"  itself 
or  the  sphere  of  money  and  credit  which  provides  the  means  for  spending  harbors  any 
cyclical  mechanism  of  its  own.  If  our  analytic  schema  be  accepted,  all  the  features  of  the 
behavior  of  spending  and  of  credit  in  a  normal  cycle  can  be  accounted  for  without  assigning 
to  them  any  but  an  adaptive  role.  Since  the  above  contains  our  solution  of  a  much-debated 
problem — it  might  be  termed  "the  riddle  of  spending" — and  since  the  opposite  view  is,  in 
one  form  or  another,  almost  universally  accepted  and  not  without  influence  on  policy,  the, 
reader  is  invited  to  recall  once  more  the  premises  of  the  statements  in  the  text. 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES       553 

not,  they  are  "maladjustments,"  which  may  have  to  be  liquidated  sub- 
sequently. Similarly,  system  expenditure  slackens  or  may  even  abso- 
lutely decrease  in  recession,  and  this  not  only  constitutes  one  of  the 
most  regular  features  of  recessions  but  also  produces  others.  But  it  is 
itself  induced  by  the  slackening  of  innovating  activity  and  by  autodefla- 
tion,  both  of  which  again  are  independent  of  it  in  the  sense  that  they 
come  about  without  spending  or  credit  taking  the  initiative  in  calling  a 
halt.  No  doubt  there  may  be,  in  any  particular  situation,  also  a  pull 
of  the  monetary  bridle.  But  (see  Chap.  IX)  such  a  pull  is  not  necessary 
in  order  to  account  for  the  occurrence  of  an  "upper  turning  point,"  which 
is  perfectly  understandable  without  it  and  does  not,  in  system  expenditure 
any  more  than  in  anything  else,  constitute  a  distinct  problem  calling  for 
special  solution  and  requiring  the  introduction  of  facts  extraneous  to  the 
mechanism  of  innovation.  In  the  midst  of  increasing  rates  of  expendi- 
ture, and  in  the  absence  of  all  extraneous  limits — monetary  or  other — 
business  could  and  would  recurrently  start  deflating  itself  by  virtue  of  the 
working  of  that  mechanism,  and  this  indeed  means  decrease  in  rates  of 
increase  or  absolute  fall  of  expenditure,  but  does  not  presuppose  that  it 
occur  from  other  reasons.  Failure  to  realize  this  leads  to  many  blind 
alleys. 

Expectation  for  the  cyclical  behavior  of  system  expenditure  seems  to 
follow  arithmetically  from  the  behavior  of  price  level  and  output. 
Expenditure  is,  however,  not  only  a  monetary  expression,  but  also  a 
monetary  quantity  that  exists  and  behaves  as  such,  so  that  our  expecta- 
tion about  it  must  be  formed  independently.  In  the  "pure"  model  it 
would  rise  in  prosperity  and  fall  to  its  previous  amount  during  liquidation. 
In  the  four-phase  cycle,  expectation  for  three  out  of  the  four  phases  is  too 
obvious  to  detain  us,  but  for  recession  it  is  necessary  to  recall  that  owing 
to  the  facts,  first,  that  entrepreneurs*  repayments  do  not  actually  go  to 
the  length  of  eliminating  their  debts  and,  second,  that  other  borrowing 
partly,  wholly,  or  more  than  wholly1  replaces  entrepreneurs'  borrowing, 
our  expectation  loses  its  definiteness.  All  we  can  say  is  that  system 
expenditure  will  increase  in  prosperity  more  and  in  recession  less  than 
total  output,  although  we  may  also  hazard  the  guess  that  it  will  increase 
in  the  recession  of  every  cycle  at  a  smaller  rate  than — due  attention  being 
paid  to  the  simultaneous  phases  of  the  other  cycles — in  the  preceding 
prosperity.  The  resulting  trend  is,  to  be  sure,  still  due  to  the  working  of 
our  process,  but  it  nevertheless  differs  in  nature  from  the  result  trends 
which  we  observe  in  price  level  or  output.  These  express  fundamental 
properties  of  the  capitalist  system  and  would  be  present  even  in  a  world 

1  That  will  depend  on  the  speed  of  the  process  of  repayment  and  of  the  process  of  exploi- 
tation of  the  new  investment  opportunities  opened  up  by  the  innovations  of  the  preceding 
prosperity,  and  on  how  this  "pushing  into  new  economic  space"  is  being  financed.  This 
is  not  without  some  suggestions  as  to  regulative  and  remedial  policy. 


554 


BUSINESS  CYCLES 


conforming  to  the  pure  model,  while  the  result  trend  in  system  expendi- 
ture would  be  absent  in  such  a  world  and  enters  our  picture  only  at  the 
level  of  the  second  approximation. 

Such  as  it  is,  that  result  trend  is  of  course  but  imperfectly  rendered  by 
the  descriptive  trend  observable  in  our  graph.  The  development  of 
deposit  banking,  the  change  in  habits  of  payment  incident  thereto,  and 
all  the  changes  in  the  institutional  framework  of  the  currency  and  in  the 


0.90 

080 

% 

5  0.70 

>° 

10 

-0.50 
c 

I  040 
0.30 
020 

£+10 


I  -10 


/  TREND  FITTED  BY  LEAST  SQUARES: 
2>0.6l747  +  O.OI780658!3t 
(ORIGIN  I896t) 


^LOGARITHMS  OF 
HILADELPHIA  CLEARINGS 
(ANNUAL  TOTALS) 
I 


DEVIATIONS  FROM 


IDEAL  FORM  OF  COMPUTED  PERIODS: 
(12.9  YEARS  AND  4  2  YEARS) 


\m  1880  1882  1884  1886  1888  1890  1892   1894  1896  1898  1900  1902  1904  1906  1908  1910   1912   1914 
CHART  XXIV.— (See  Appendix,  p.  1060.) 

rate  of  gold  production  also  help  to  shape  the  latter.  It  must  be  recog- 
nized, moreover,  that  the  effects  of  all  these  factors  are  clearly  not  addi- 
tive and  that  the  effects  of  changes  in  the  quantity  of  legal  tender,  in 
particular,  cannot  be  dealt  with  in  the  way  which  would  have  seemed — 
and  still  seems — the  obvious  one  to  adherents  of  the  quantity  theory  in 
any  of  its  primitive  versions.  The  circumstance  that  none  of  those  fac- 
tors is  independent  of  our  process  may  perhaps  be  pointed  to  in  explana- 
tion of  the  fact  that,  barring  times  of  war  inflation,  the  behavior  of  our 
series  is,  nevertheless,  quite  satisfactorily  accounted  for  by  our  model  and 
that  it  does  not  obviously  indicate  any  other  than  a  passive  role  of  money 
and  credit. 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES      555 

Chart  XXIV  is  presented  for  the  sake  of  displaying  an  ingenious 
method  rather  than  for  the  results  that  can  be  gathered  from  it.  In  order 
to  apply  Dr.  Georgescu's  method  (Chap.  V),  it  was  necessary  (the  other 
two  differentiations  which  would  have  had  to  be  added  would  have 
accumulated  errors  to  a  dangerous  extent)  to  eliminate  a  descriptive  trend 
fitted  by  least  squares.  The  reader  is  invited  to  record  and  weigh  this  fact 
against  what  has  been  said1  about  this  practice  in  the  course  of  our  con- 
siderations on  statistical  method  (Chap.  V).  Moreover,  the  material 
covers  too  short  a  span  to  allow  any  formal  method  to  show  the  Kondra- 
tieff,  or  even  to  give  reliable  results  as  to  the  precise  periods,  or  any  precise 
properties,  of  the  shorter  cycles.  However,  the  hypothesis  of  a  single 
cycle  yielded  a  period  of  10M  years,  which,  the  writer  holds,  is  strong 
evidence  both  of  the  presence  of  a  cycle  of  about  that  length  and  also  of 
its  general  properties'  conforming  to  our  expectations.  The  second  step 
of  this  analysis  turned  out  less  satisfactorily.  When  the  material  was 
analyzed  on  the  assumption  of  the  presence  of  two  cycles,  the  resulting 
periods  were  12.9  and  4.2  years  respectively.  We  should  not  take  it  too 
tragically  that  these  figures  exceed  substantially  what  we  consider  to  be 
the  average  periods  of  the  Juglars  and  the  Kitchins.  The  fewness  of  the 
cyclical  units  of  both  kinds  may  account  for  that,  and  the  value  of  the 
evidence  for  the  presence  of  two  cycles  of  about  these  periods  is  not 
greatly  impaired  thereby.  But  it  is  more  serious  that  the  introduction 
of  the  hypothesis  of  two  cycles  does  not  improve  the  picture  of  the 
phenomenon  yielded  by  the  single-cycle  hypothesis.  The  "fit"  is  some- 
what improved,  but  not  much.2 

Turning  to  Chart  XXV,  we  will  first  of  all  note  the  fact  adverted 
to  above  that  not  only  the  properties  of  the  clearings  series  to  be 
expected  from  our  model  do  actually  show,  but  also  that  it  behaves  much 
as  if  no  other  influences  than  those  embodied  in  our  model  had  acted 
upon  it.  For  1875  to  1913  this  is  actually  more  nearly  true  than  for  any 
other  period  within  our  range  of  vision.  But  we  should  not  have  been 
surprised  if  such  outside  disturbances  as  nevertheless  occurred  had 
asserted  themselves  more  clearly.  Detailed  analysis  no  doubt  reveals 
some  of  them.  The  steepness  of  the  ascent  from  1878  has,  for  example, 
perhaps  something  to  do  with  the  certainty  of  "sound  money"  and 
affords  in  any  case  an  interesting  lesson  about  how  business  reacts  in  the 
face  of  what  many  economists  would  describe  as  deflationary  tendencies. 
The  ascent  from  1897  may  have  something  to  do  with  gold  or,  more 

1  The  reader  will  reflect  however  that,  in  statistics  as  in  life,  the  argument  against  Sin 
is  not,  in  good  logic,  weakened  by  the  observation  that  the  preacher  sins  himself. 

2  The  results  of  the  two  tests  carried  out  are  not  shown  in  the  graph.     It  should  be  men- 
tioned, that  United  States  and  English  clearings  were  among  the  series  by  the  analysis  of 
which  Mr.  Kitchin  first  showed  the  existence  of  the  cycle  which  we  refer  to  by  his  name.    He 
also  showed  that  "two  or  three"  of  them  seem  to  form  a  higher  unit  (our  Juglar). 


556 


BUSINESS  CYCLES 


plausibly,  with  the  practices  of  trust  companies,  but  would  be  under- 
standable without  either.  Nothing  indicates  obviously  the  presence 
of  other  than  our  evolutionary  factors. 

The  long  wave  could  not,  in  the  absence  of  other  information,  be 
discerned  in  this  graph  any  more  than  in  Chart  XXIV,  since  our  series  covers 
only  the  end  of  the  second  (1875  to  1897)  and  the  beginning  of  the  third 
Kondratieff.  But  if  we  tried  to  fit,  by  the  method  of  least  squares, 


LOANS  AND  DISCOUNTS 
OUTSIDE  N.Y.C. 


1875         I860         1885        1890         1895        1900        1905        1910   1913 
CHART  XXV. — United  States  (see  Appendix,  p.  1000). 

either  a  straight  line  or  a  second-degree  parabola,  we  should  discover, 
as  others  have  before  us,  that  for  some  reason  a  "break  in  trend"  occurred 
in  the  nineties.  If  thereupon  we  proceeded  to  fit  two  straight  lines, 
the  one  to  the  interval  from  1875  to  1897  and  the  other  to  the  interval 
from  1898  to  1913,  we  would  find  that  the  gradient  of  the  former  is 
smaller  than  the  gradient  of  the  latter.  It  is  submitted  that,  since  we 
know  about  the  Kondratieffs  from  other  evidence,  historical  as  well  as 
statistical,  and  since  that  behavior  of  clearings  conforms  to  what  we 
should  expect  it  to  be  in  those  intervals,  there  is  some  point  in  inter- 
preting it  as  a  Kondratieff  effect.  This  would  not  only  solve  the  problem 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES       557 

of  that  break  in  trend,  but  also  enable  us  to  account  for  the  fundamental 
contours  of  United  States  clearings  in  that  period  by  a  single  set  of  prin- 
ciples. Whether  this  is  more  convincing  than  an  explanation  in  terms  of 
gold  production,  is  for  the  reader  to  decide. 

The  shorter  cycles  stand  out  well  and  show  in  the  two  intervals  in 
the  way  we  should  expect  them  to  show  in  Kondratieff  depressions, 
revivals,  and  prosperities.  The  series  is,  in  fact,  extremely  "responsive." * 
We  behold  at  the  beginning  the  shrinkage  of  system  expenditure  incident 
to  the  deep  depression  of  the  seventies.  Then  follow  two  very  well- 
marked  Juglars  (1879-1888,  1889-1897)  within  each  of  which  three 
Kitchins  are  clearly  discernible.  The  strong  swell  of  the  prosperity  of 
the  third  Kondratieff  dominates  the  picture  after  1897  and  tends  to  iron 
out  the  other  two  cycles,  which  assert  themselves  nevertheless.  The 
relation  of  the  clearings  series  to  others  follows  from  what  has  been  said 
before  in  this  and  preceding  chapters.  As  regards  timing,  the  reader 
should  again  be  warned  not  to  form  rash  expectations  about  consistent 
sequences  or  to  draw  inferences  from  a  failure  to  find  such  sequences. 
For  instance,  he  may  feel  tempted,  considering  that  system  expenditure  is 
intimately  related  to  the  initiating  impulse  of  the  cyclical  process  and 
that  it  is  itself  the  immediate  source  of  many  of  the  symptoms  of  general 
prosperity  or  depression,  to  expect  that  variations  of  clearings  should 
precede  variations  in  most  other  cyclical  quantities  and,  especially, 
variations  in  price  level.  As  a  matter  of  fact,  they  mostly  do,  within 
short -time  fluctuations,  by  a  few  months.  But  even  apart  from  internal 
irregularities,  the  influence  of  speculative  anticipation,  the  fact  that 
prices  are  made  by  contracts  while  clearings  reflect  actual  payment,  and 
so  on,  there  is  no  theoretical  reason  for  expecting  precedence  of  clearings 
for  all  phases  of  a  cycle  but  only  for  the  positive  ones,  and  for  these  only 
with  a  proviso  as  to  the  simultaneous  phases  of  the  other  cycles.  Statis- 
tical measurement  of  covariation  by  means  of  formal  methods  is  bound 
to  give  disappointing  and  inconclusive  results,2  particularly  if  the  method 
used  implies  the  single-cycle  hypothesis. 

Much  more  interesting  than  variations  of  total  system  expenditure, 
are  the  variations  of  its  two  principal  constituents — consumers'  and 
producers'  expenditure.  But  even  for  postwar  times  we  are  far  from 
having  adequate  data  from  which  to  compile  separate  indices  of  them, 
while  for  prewar  times  we  must  not  be  under  any  delusion  as  to  the  value 

1  Of.  Professor  Crum's  Interpretation  of  the  Index  of  General  Business  Conditions, 
Review  of  Economic  Statistics,  Supplement  2,  September  1925. 

2  The  Harvard  Committee's  correlation  between  outside  clearings  and  Bureau  of  Labor 
Statistics  prices  at  wholesale  (both  corrected  for  seasonal  variation  and  trend),  for  1903- 
1918,  is  described  as  Fair  with  a  four  months'  lead  of  poor  consistency  in  the  former 
(Review  of  Economic  Statistics,  1919,  p.  184). 


558 


BUSINESS  CYCLES 


of  any  inference  we  may  draw  from  such  material  as  we  have.  Efforts  in 
that  direction  are  nevertheless  among  the  most  urgent  desiderata  econo- 
mists have  to  address  to  public  and  private  organizations  commanding 
the  necessary  means.  What  we  can  do  here  falls  very  far  short  of  what 
might  be  accomplished  even  now. 


A 

f  \ 


EXPENDITURE  ON   / 
PRODUCERS*  GOODS/ 


EXPENDITURE  ON 
CONSUMERS' GOODS, 


V- 

A 


"'OUTSIDE 
CLEARINGS 


PRODUCERS'  GOODS 
EXPENDITURE 

OUTSIDE  CLEARINGS 
A 


\    : 

V 


CONSUMERS*  GOODS 

EXPENDITURE 
OUTSIDE  CLEAR  NGS 


1875         I860         1885          1890         1895         1900          1905         1910    1913 
CHART  XXVI.— United  States  (see  Appendix,  p.  1060). 

A  rough  indication  of  the  behavior  of  consumers'  expenditure,  the 
most  important  figure  of  monetary  theory,  may  be  derived  from  the  series 
of  total  wages  disbursed  as  represented  by  pay  rolls,  since  this  is  not 
only  the  biggest  individual  item  in  the  sum  total  of  incomes,  but  also 
the  one  which  may  be  expected  to  bear  a  closer  relation  to  actual  expendi- 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES      559 

ture  on  consumers'  goods  than  any  other.  In  spite  of  all  qualifications 
which  have  to  be  made  but  are  too  obvious  to  warrant  restatement,  we 
may  again  compare  wage  bill  and  outside  clearings  (Chart  XXIII). 
The  line  at  the  bottom  of  the  chart  will  help  in  interpreting  the  result. 
The  sharp  rise  of  the  clearings-pay  roll  ratio  at  the  beginning  of  the 
current  Kondratieff  is  particularly  instructive.  Since  the  question  of 
how  the  sum  total  of  taxable  incomes  is  related  to  consumers'  expenditure 
is  controversial,  we  will  not  press  it  into  the  same  service. 

Another  indication  of  the  behavior  of  the  two  main  constituents  of 
system  expenditure  is  given  (Chart  XXVI)  by  the  index  of  physical 
quantity  of  consumers'  goods  times  the  index  of  consumers'  goods'  prices 
and  the  index  of  physical  quantity  of  producers'  goods  times  the  index 
of  producers'  goods'  prices.  Our  chart  is  unsatisfactory  in  many  respects, 
and  no  confidence  can  be  placed  in  the  details  of  the  picture.  But  how- 
ever much  progress  in  the  direction  of  exact  measurement  we  may  expect 
from  more  thoroughgoing  investigation  which  government  agencies  or 
research  groups  might  undertake,  it  is  extremely  unlikely  that  it  would 
result  in  substantially  different  contours  or  that  errors  should  have  sys- 
tematically worked  so  as  to  give  spurious  support  to  our  theoretical 
structure. 

We  notice,  first,  that  Juglars  and  Kitchins  show  well  in  both  the 
dollar  values  of  consumers'  and  in  the  dollar  values  of  producers'  goods, 
which  we  assume  to  indicate  consumers'  and  producers'  expenditure. 
In  the  case  of  the  Kitchins  it  is,  no  doubt,  necessary  to  recollect  what  has 
on  previous  occasions  been  said  about  the  likelihood  of  their  being  "ironed 
out"  in  very  strongly  marked  phases  of  the  longer  cycles,  but  they 
remain  always  discernible,  at  least  in  rates.  The  rise  of  the  third  Kondra- 
tieff, 1898  to  1907,  is  impressively  reflected.  Second,  we  observe  the 
difference  in  amplitudes  of  fluctuations,  a  familiar  feature,  which  need 
not  detain  us.  Third,  it  is  seen  at  a  glance  that  producers'  expenditure 
dominates  the  contour  of  the  clearing  series.  Fourth,  producers'  expendi- 
ture can  on  the  whole  be  said  to  precede,  although  not  consistently. 
For  reasons  repeatedly  mentioned  we  would  attach  little  importance  to 
this,  even  if  our  chart  were  as  perfectly  exact  as  it  is  defective.  We  have 
seen  reason  to  question  the  validity  of  causal  interpretation  of  statistical 
sequences;  theoretically  we  cannot  with  any  confidence  expect  that  pro- 
ducers' expenditure  should  precede  in  time  as  well  as  in  logic,  for,  effects 
on  consumers'  expenditure  asserting  themselves  quickly,  we  might  well 
find  roughly  simultaneous  covariation  blurred  by  secondary  and  random 
influences  rather  than  a  consistent  sequence  conforming  to  that  logic; 
and,  finally,  the  case  well  exemplifies  the  possibility  that  a  "logical" 
sequence  be  inverted  by  anticipation,  for  even  if  consumers'  expenditure 
were  the  prime  mover  of  the  process,  producers'  anticipations  might 


560  BUSINESS  CYCLES 

easily  make  producers'  expenditure  move  first.1  Without,  however, 
receding  from  this  standpoint,  we  may  still  notice  a  possibly  significant 
fact.  If  the  reader  smooth  both  curves  by  freehand,  he  will  get  a  picture 
that  roughly  renders  the  Juglars  and  the  Kondratieff  effect.  And 
within  this  picture,  precedence  of  producers'  expenditure  is  unmistakable 
and  consistent.  Inconsistencies  of  lag  may  thus  be  shown  to  be  due 
to  the  shorter  fluctuations.  Since  these  are  particularly  exposed  to 
disturbing  influences  and  fundamental  relations  work  ^  themselves  out 
more  clearly  in  the  longer  cycles,  that  fact  may  mean  something,  after  all. 

Now,  these  four  features  certainly  tend  to  justify — no  proof  is  ever 
possible  by  statistics  alone — our  decision  to  consider  producers'  expendi- 
ture as  the  "active"  element  in  total  expenditure.  We  move  on  common 
and  familiar  ground  when  we  say  that  expenditure  on  plant  and  equip- 
ment is  the  "  active  "  factor  in  producers'  expenditure.  But  if  we  go  on  to 
say  that  expenditure  on  the  creation  of  new  production  functions — 
innovation — is  the  active  element  in  total  expenditure  on  plant  and 
equipment,  we  are  leaving  this  terra  firma,  and  there  is  every  reason  for 
the  reader  to  examine  the  nature  of  this  third  step.  It  is  not,  like  the 
two  others,  a  formulation,  but  an  interpretation  of  time-series  fact.  It 
adds  an  explanatory  hypothesis.  But  this  hypothesis  justifies  itself, 
by  solving  a  problem  presented  by  time-series  fact — namely,  the  problem 
of  the  wavelike  character  of  expenditure  on  plant  and  equipment — and 
by  doing  this,  through  its  appeal  to  intermittent  entrepreneurial  impulses, 
in  a  way  that  leads  us  out  of  the  circle  of  perpetuum  mobile  theories.2 
Moreover,  it  follows  from  a  schema  all  the  expectations  from  which 
are  borne  out  by  time-series  fact,  where  such  fact  is  available.  Finally, 
it  is  fully  verified  by  economic  history:  we  simply  know  what  innovations 
were  responsible  for  the  humps  in  the  curve  of  values  of  producers'  goods 
from  1879  to  1882,  in  the  late  eighties,  and  in  the  late  nineties,  or  at  the 
beginning  of  the  century. 

Attention  is  called  to  the  two  lines  at  the  bottom  of  Chart  XXVI, 
which  present  the  fluctuations  of  consumers'  and  producers'  expenditure 
per  dollar  of  system  expenditure.  As  we  might  expect,  they  demonstrate 

1  Compare  again  the  discussion  between  Professors  Frisch,   Hansen,  and  Clark  on 
Capital  Production  and  Consumer-taking,  Journal  of  Political  Economy,  1921-1922. 

2  The  problem  may  also  be  formulated  as  follows:  How  is  it  possible  for  prosperity  to 
start,  let  us  say,  from  perfect  equilibrium,  without  an  impulse  being  first  given  to  produc- 
tive activity  by  consumer's  expenditure?     It  would,  indeed,  be  impossible  but  for  inno- 
vation's breaking  through  the  existing  system  of  production  functions  and  creating  an 
opening  for  further  outlay — new  profitable  opportunities  for  the  expansion  of  producers' 
expenditure,  or,  as  we  may  call  it,  new  "investment."     Hence  the  mere  fact  of  those 
recurring  bursts  of  "investment'*  might  be  held  to  furnish  verification  of  the  presence  of 
what  is,  unless  they  are  explained  by  external  factors,  their  only  adequate  en  use,  the 
element  that  induces  all  others. 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES       561 

with  particular  emphasis  what  has  just  been  explained.  Both  the  rhythm 
and  the  mechanism  of  our  process — in  particular,  the  character  of  pros- 
perities as  periods  of  investment  and  the  character  of  recessions  and 
revivals  as  periods  of  consumers'  harvests — assert  themselves  very 
instructively.  But  there  is  also  another  movement — the  long-time  and 
systematic  change  in  the  relation  between  consumers'  and  producers' 
expenditure,  which  we  already  have  had  occasion  to  observe  in  the  realm 
of  physical  quantities,  and  which  lends  itself  so  readily  to  interpretation 
in  the  sense  of  Boehm-Bawerk's  theory.  The  fact  that  investment  gains 
on  consumers'  expenditure  in  prosperities  without  ever  losing  ground  in 
the  long  run,  which  stands  out  particularly  for  the  period  of  the  Kondra- 
tieff  upswing,  accounts  for  the  only  long-run  difference  there  is  between 
the  behavior  of  producers'  expenditure  and  clearings.  Outside  Clearings 
in  turn  should  now  be  compared  with  outside  deposits  and  loans;  but 
before  taking  up  this  matter,  we  will  digress,  in  order  to  discuss  the 
behavior  of  total  income  and  of  wages. 

C.  National  Income  and  Wages. — 1.  The  familiar  difficulties  which 
we  all  experience  in  defining  National  Income  are  of  course  due  to  the 
fact  that  it  is  not  a  technical  term  wedded  to  one  definite  use  but  a  word  of 
common  parlance  that  is  loosely  used  for  a  great  many  purposes  which 
cannot  be  served  equally  well  by  a  single  definition.  For  our  purpose 
we  need  not  discuss  the  general  merits  or  demerits  of  inclusion  or  exclusion 
of  annual  services  of  durable  goods  or  of  the  values  of  services  rendered  by 
the  individual  household  to  itself  or  of  the  values  of  receipts  in  kind  or  of 
the  gains  from  cooperative  purchasing,  or  any  of  the  other  questions  that 
arise  in  defining  national  income  and  in  evaluating  it  statistically, 
although  some  of  them  will  have  presently  to  be  mentioned  for  the  special 
case  of  wages.  At  the  moment,  we  will  simply  think  of  a  monetary 
quantity  (hence,  not  identical  with  value  of  output)  that  consists  of  the 
receipts  of  households  from  the  sale  of  services,  personal  or  other — sub- 
stantially wages,  rents,  quasi-rents — plus  profits  and  interest,  but 
exclusive  of  gains  from  the  sale  of  capital  assets,  appreciation  of  inven- 
tories, and  so  on.  This  total  is  equal  to  consumers'  expenditure  (on 
durable  as  well  as  on  transient  goods),  households'  investments  (for 
definition  see  Chap.  Ill,  Sec.  A),  and  tax  payments,  minus  households' 
net  borrowings  and  expenditure  of  capital  gains,  plus  sums  that  are  not 
spent  at  all  (which  we  know  can  occur  equally  well  in  the  cases  of  sums 
earmarked  for  consumption,  as  in  the  cases  of  sums  earmarked  for 
investment),  and  it  makes,  for  most  purposes,  a  highly  inconvenient 
composite.  But  it  serves  for  our  purpose. 

The  corresponding  time  series  is  obviously  natural,  primary,  conse- 
quential, systematic,  and  cyclical.  It  displays  a  result  trend  in  the  same 
sense  as  system  expenditure,  that  is  to  say,  only  because  and  as  far  as  the 


562  BUSINESS  CYCLES 

cyclical  process  of  evolution  expands  the  elastic  strings  of  the  monetary 
ligamina  for  good.  Barring  this,  there  would  be  no  result  trend  in  it, 
and  national  income  in  that  sense  would,  in  the  process  described  by  the 
pure  model,  return  in  each  neighborhood  to  its  value  in  the  preceding  one, 
which  it  is  not  superfluous  to  mention  in  view  of  an  application  to  be 
made  of  this  proposition  in  the  theory  of  the  cyclical  behavior  of  the  wage 
bill.1  In  every  four-phase  cycle  we  shall  still  expect,  with  the  usual 
qualification  about  the  simultaneous  phases  of  the  other  cycles,  that  our 
total  increases  in  prosperity  more  than  does  the  output  of  consumers' 
goods.  But  expectation  as  to  its  behavior  in  recession  is  uncertain 
beyond  two  points :  we  shall  predict  that  if  it  increases  at  all,  it  will  do  so 
at  a  rate  that,  on  the  one  hand,  is  smaller  than  in  the  preceding  prosperity 
and,  on  the  other  hand,  is  smaller  than  the  rate  of  increase  in  output  of 
consumers'  goods.  In  depression  it  will,  in  general,  fall — at  least  in 
"deep"  depression — and  in  recovery  it  will  return  to  equilibrium  amount. 
The  only  data  at  all  adequate  for  the  purpose  of  compiling  a  series 
of  national  income,  are,  for  prewar  times  at  least,  those  supplied  by  the 
materials  of  the  English  and  the  Prussian  income  taxes.2  The  first, 
covering  the  whole  period  since  Sir  R.  Peel's  rein troduct ion  of  the 
income  tax,  is  the  one  we  shall  use,  because  the  second,  the  figures  of 
which  should,  owing  to  the  superior  technique  of  Miquel's  income  tax  act, 
really  be  more  valuable,  covers  only  a  little  more  than  the  years  of  the 
third  Kondratieff.3  Of  course,  the  series  of  taxable  income  is  not  what 
we  want  and  can  only  in  a  very  rough  way  indicate  the  movement  of  the 
quantity  we  are  interested  in.  Taxable  income  covers  both  less  and  more. 
Since  the  wage  bill  or,  more  correctly,  that  part  of  the  sum  total  of  wages 

1  Strictly  speaking,  it  would,  for  that  proposition  to  be  borne  out  by  statistical  fact, 
also  be  necessary  that  the  sums  saved  and  the  sums  paid  in  taxes  from  income  were  equal 
in  both  neighborhoods.  For  instance,  an  increase  of  the  income  tax  for  the  purpose  of 
increasing  the  salaries  of  public  servants  would  increase  national  income  as  above  defined 
and  an  analogous  proposition  would  hold  true  for  savings.  This  is  one  of  the  reasons  why 
this  definition  has  been  called  inconvenient  in  the  text:  a  definition  that  entails  the  conse- 
quence that  national  income  falls  if  the  sum  total  of  net  saving  falls  has  certainly  no  claim 
to  general  acceptance. 

a  There  is,  for  the  United  States,  an  estimate  by  the  Brookings  Institution  of  national 
income  paid  out,  for  1901  and  for  1910  to  1914;  and  by  the  Cleveland  Trust  Company  for 
1902  to  1909  (see  Bulletin  of  the  Cleveland  Trust  Company  for  Apr.  15,  1935).  See  also, 
W.  I.  King's  National  Income  and  Its  Purchasing  Power,  1930.  But  in  an  extremely 
rough  approximation,  railroad  gross  earnings  may  perhaps  give  an  index;  see  A.  H.  Cole, 
A  Monthly  Index  of  Railroad  Earnings,  1866-1914,  chart  comparing  this  index  with 
Partington's  quarterly  series,  on  p.  41,  Review  of  Economic  Statistics  for  February  1936. 
Professor  Cole's  index  is  reproduced  on  Chart  XXIX. 

8  There  were  other  income  taxes  in  German  states,  some  of  which — for  instance,  the 
one  of  Saxony  and  the  one  of  Baden — give  useful  indications.  See  a  study  on  Einkommen- 
schwankungen  vor  dem  EJriege  by  the  Institut  ftir  Konjunkturforschung,  Vierteljahrshefte, 
1927.  Results  do  not  differ  from  those  that  we  are  about  to  derive  from  the  English  series. 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES      563 

that  is  not  subject  to  the  tax,  could  be  added  from  1860  on,  and  since  that 
part  of  "unearned"  income  that  goes  to  income  receivers  below  the 
exemption  limit  may  be  estimated  from  the  data  of  the  Inland  Revenue 
reports,  this  is  less  serious  than  the  effect  of  exemptions,  allowances, 
and  abatements,  on  the  one  hand,  and  inclusion  of  various  items  that 
ought  not  to  figure,  on  the  other.1  It  must  also  be  borne  in  mind,  of 


1780     1790 


1800      1810       1820       1830      1840      1850      1860      1870      I860      1890      1900      191013 
CHART  XXVII. — United  Kingdom  (see  Appendix,  p.  1060). 


course,  that  income  from  foreign  sources  cannot  be  expected  to  vary  as 
home-produced  income. 

The  picture  presented  in  Chart  XXVII  cannot  hence  be  implicitly 
trusted.  It  is,  nevertheless,  not  very  hazardous  to  say  that  it  bears  out 
expectations  as  to  Juglars  in  which  total  taxable  income  in  fact  displays 
a  tendency  to  rise  in  prosperities  and  to  stay  at  about  the  figure  reached, 
or  very  little  less,  for  the  rest  of  each  cycle.  The  reader  can  easily  satisfy 
himself  of  this  by  recalling  the  dates  of  the  Juglar  prosperities.  Since 
the  technique  of  the  assessment  of  business  profits  has  a  smoothing  effect, 
we  shall  understand  that  we  see  little  of  a  Kitchin  movement,  but  the 
behavior  of  the  series  in  the  Kondratieff  wave  is  contrary  to  expectation 
up  to  1873. 2  From  that  year  on  until  1898  we  observe  a  remarkable,  if 

1  Much  work  has  been  done  in  order  to  improve  the  material  and  to  distill  from  it 
really  relevant  figures.     See,  in  particular,  Sir  G.  Paish's  paper  in  the  Journal  of  the  Royal 
Statistical  Society,  1909;  and  Sir  J.  Stamp's  British  Incomes  and  Property. 

2  The  strong  increase  of  total  taxable  income  from  1860  to  1873  indicates  the  presence 


564  BUSINESS  CYCLES 

rough,  parallelism  of  "trend"  with  the  production  series,  while  there  is 
none  with  the  series  of  production  times  price  level.  From  about  1876 
to  1913  parallelism  with  the  wage  bill  is  no  less  remarkable.  This  fact 
is  in  itself  sufficient  to  dispose  of  some  of  the  more  primitive  errors  about 
the  relation  of  wages  and  other  incomes.  The  relation  to  provincial 
clearings  is  what  we  should  expect  it  to  be. 

The  behavior  of  the  series  of  total  taxable  income  is  obviously  domi- 
nated by  business  profits  in  the  usual  sense,  as  the  reader  may  see  by 
referring  to  Chart  XXVIII.  The  " profits"  there  charted  are  simply 
from  the  data  of  Schedule  D,  which  include  not  only  interest,  royalties, 
and  the  like,  but  also  a  large  amount  of  income  that  ought  to  be  classed 
with  wages.  The  distinction  made  according  to  formal  rules  between 
" earned"  and  " unearned"  income  does  not,  of  course,  help  us  at  all. 
We  may  look  upon  that  item  as  roughly  rendering  the  income  from 
"business."  But  it  is  far  removed  from  profits  in  our  sense,  which  form 
an  unknown  and  varying  fraction  of  it.  It  is  again  by  way  of  an  inter- 
pretation, which  is  not  gleaned  from,  but  added  to,  statistical  fact,  that 
it  can  be  asserted  that  entrepreneurial  activity  is  what  directly  and 
indirectly  accounts  for  the  waves  we  observe.  For  Germany  we  have 
also  a  fairly  long  series  of  dividends  (per  cents  of  nominal  capital),  see 
Chart  XXXVIII.  Dividends  are  of  course  not  profits — still  less  are  they 
profits  in  our  sense — but  the  relation  is  close  enough  to  permit  taking  them 
as  an  indication.  Perhaps  we  need  not  stress  the  proviso  usually  made 
about  lag,  for  the  state  of  affairs  at  the  moment  of  the  decision  about 
dividends  in  many  cases  influences  that  decision  as  much  as  the  results 
achieved  in  the  preceding  business  year  do.  On  the  other  hand,  the 
policy  of  equalizing  returns  to  shareholders,  which  prevails  in  other 
cases,  naturally  dampens  fluctuations. 

2.  We  must  go  more  fully  into  the  behavior  of  wages.  There  is  a 
considerable  amount  of  information  for  earlier  times,  sometimes  per- 
mitting alignment  into  time  series,  practically  always  enabling  us,  in 
spite  of  all  difficulties  about  currency  and  prices,  to  discern  the  roughest 
contour  lines,  at  least  for  Germany  and  Western  Europe.1  By  them- 

of  a  disturbance  external  to  our  model.  Since  the  anomaly  obviously  links  up  with  the 
anomaly  we  observed  in  the  price  level  series,  we  may  attribute  it  to  gold.  When  gold 
production  turned  down,  total  taxable  income  assumed  what  we  could  conceive  to  be  its 
normal  behavior. 

1  E.  J.  Hamilton,  Money  Prices  and  Wages  in  Valencia,  Aragon,  and  Navarre,  1859- 
1500,  1986;  and  Prices  and  Wages  at  Paris  under  John  Law's  System,  Quarterly  Journal  of 
Economics  for  November  1986.  See  also,  besides  the  familiar  French  and  English  works  on 
the  subject,  such  as  Thorold  Rogers'  or  Levasseur's,  E.  W.  Gilboy,  Wages  in  Eighteenth 
Century  England,  1934;  and  Schmoller's  handy  survey  (antiquated  in  part,  of  course), 
Tatsachen  der  Lohnbewegung,  in  his  Jahrbuch  1914,  republished  in  the  second  edition  of 
his  Grundriss,  particularly  valuable  because  of  the  insistence  on  the  institutional  setting, 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES       565 

selves,  of  course,  those  data  mean  very  little,  particularly  with  respect  to 
any  welfare  considerations.  Recent  work  has  done  much  to  enlighten 
us  about  the  eighteenth  century,  notably  in  England,  and  more  may  be 
hoped  for  in  the  near  future.  But  at  every  step  so  far,  we  find  room  for 
difference  of  opinion  as  to  questions  of  fact.1  Mrs.  Gilboy's  book  brings 
out  sharply  that  even  England  was  then  not  yet,  with  respect  to  wages, 
an  economic  domain  in  the  theorist's  sense  and  that  wage  rates  at  times 
even  moved  in  different  directions  in  different  parts  of  the  country.  We 
do  not  venture  to  go  beyond  stating  that,  on  the  whole,  money  wage  rates 
started  to  rise,  or  in  some  cases  continued  to  rise,  after  1750;  that  they 
rose  during  the  prosperity  phase  of  the  first  Kondratieff ;  while  real  wage 
rates  on  the  whole  rose  very  much  less,  if  indeed  they  rose  at  all.2  For 
the  nineteenth  century  we  rely  on  the  work  of  Mr.  Wood  and  of  Professor 
Bowley,  which,  for  agriculture  and  building,  also  includes  some  eigh- 
teenth-century data. 

For  this  country  we  have  scattered  information  for  earlier  times,  most 
of  which  is  reproduced  or  noticed  in  V.  Clark's  work.  Nation-wide 
figures  that  are  at  all  reliable  do  not  date  back  of  this  century.  We  may 
mention,  however,  Bulletin  499  of  the  United  States  Bureau  of  Labor 
Statistics,  which  attempts  to  cover  available  material  from  colonial 
times  to  1928  and  Bulletins  59,  65,  and  77,  for  the  time  from  1890;  the 
National  Industrial  Conference  Board  studies  on  the  subject,  as  well  as 
the  work  of  Mr.  Rubinow  and  Professors  Hanson  and  Wolman;  but 
especially  Professor  Paul  Douglas's  treatise  on  Real  Wages  in  the  United 
States,  1890-1926,  to  which  the  present  writer  is  mainly  indebted.  We 
are  very  badly  off  for  Germany  until  almost  the  end  of  our  period. 
There  are,  however,  more  than  a  dozen  individual  series,  of  which  the 
wages  of  miners  in  the  basin  of  the  Ruhr,  beginning  in  1850,  is  the  most 
interesting  one.3  We  have  confined  ourselves  mainly  to  the  English 

which  goes  far  toward  bringing  out  the  true  meaning  of  the  wage  data  themselves.  F. 
Simiand's  work,  Le  Salaire,  1'evolution  sociale  et  la  monnaie,  essai  de  theorie  experimentale 
(sic)  du  salaire,  1932,  much  disfigured  by  methodological  dissertations  of  doubtful  value 
and  highly  questionable  theorizing,  gives  French  wage  rates  per  day  from  1789  to  1930 
and  very  useful  documentation.  We  will  mention,  finally,  the  interesting  article  by  D. 
Knoop  and  G.  P.  Jones,  Masons'  Wages  in  Mediaeval  England,  in  Economic  History  for 
January  1933;  see  chart  on  p.  486. 

1  See,  for  example,  the  controversy  between  Mr.  Hammond  and  Professor  Clapham 
(Economic  History  Review^  1930,  and  the  latter's  reply  in  the  preface  to  the  second  edition 
of  his  volume  on  the  Early  Railroad  Age)  about  the  state  of  things  at  the  end  of  the  eight- 
eenth century  and  the  beginning  of  the  nineteenth. 

2  Much  bolder  statements  are  made  in  Mr.  R.  Tucker's  article  in  the  Journal  of  the 
American  Statistical  Association  for  March  1936.     But  see  E.  W.  Gilboy's  Cost  of  Living 
and  Real  Wages  in  Eighteenth  Century  England,  Review  of  Economic  Statistics,  1936. 

3  Two    books    may  be  mentioned-  R.   Kuczynski,   Arbeitszeit  und  Arbeitslohn;  and 
Tyszka,  Lohne  und  Lebenskosten  in  Westeuropa  im  19  Jahrhundert.     The  very  good 


566  BUSINESS  CYCLES 

case.  All  data  are  primarily  for  money  rates.  Wage  bill  totals  are  not 
available  for  most  of  our  period  and  for  most  countries.  But  there  is  an 
English  series  from  1860  on,  which  we  owe  to  Professor  Bowley.1  The 
American  figure  we  use  is  but  an  indicator  of  the  behavior  of  the  national 
pay  roll.  Moreover,  most  wage  data  fail,  even  at  their  best,  to  represent 
wages  actually  paid,  which  may  differ  quite  considerably  from  official, 
particularly  trade-union  figures,  which  sometimes  veil  a  fall  and  at  other 
times  represent  minima  only.  It  is  believed,  however,  that  most  of  the 
conclusions  of  this  section  are  outside  of  the  danger  zone  created  by  those 
and  other  defects  in  our  data. 

All  sorts  of  difficulties  arise  previously  to,  and  independently  of,  those 
that  are  inherent  to  our  material.  We  will  mention  two.  There  is,  first, 
a  difficulty  about  the  delimitation  of  the  returns  that  should  be  looked 
upon  as  wages.  This  term  means  different  things  to  the  economist  and  to 
the  sociologist,  and  the  economic  class  of  wage  earners  is  much  wider  than 
the  social  class  to  which  we  refer  by  the  same  term.  But  although  part 
of  the  salaries  and  other  emoluments  of  managers  and  executives  should, 
by  the  economist,  be  included  in  wages,  another  part  is  a  rough  contractual 
equivalent  for,  or  share  in,  profits  in  our  sense.  An  analogous  difficulty 
arises  about  the  incomes  of  independent  business  men,  while  professional 
incomes  are  almost  exclusively  wages.  In  practice,  of  course,  we  simply 
use  the  material  we  have.  Second,  excepting  series  of  wages  paid  to 
homogeneous  groups  of  workers  in  a  given  firm  or  industry  or  neighbor- 
hood, the  only  natural  series  we  have  is  the  series  of  the  money  wage  bill. 
All  others  are  synthetic  ones  and  raise,  anterior  to  questions  of  deflating, 
index  problems  incident  to  the  concept  of  the  average  wages  of  a  working 
population.2 

We  distinguish  the  Sum  Total  of  Wages  (or  Payroll  or  Wage  Bill), 
Wage  Rates  (per  unit  of  time  or  product  or  per  head  of  population 
employed  or  of  population  seeking  employment,  ideally,  per  man-hour) 

Swedish  data  are,  again,  in  a  class  by  themselves;  see  G.  Bagge,  Wages  in  Sweden  1830- 
1980,  vol.  II  of  the  Stockholm  Economic  Studies. 

1  Tests  of  National  Progress,  Economic  Journal  for  September  1904,  continued  by  Pro- 
fessor Pigou,  see  Table  III  of  the  appendix  to  Industrial  Fluctuations.     Professor  Bowley, 
in  kindly  permitting  the  writer  to  use  those  figures,  expressed  a  wish  that  it  should  be  stated 
that  he  has  revised  his  treatment  in  a  book  which  has  since  been  published  by  the  Cam- 
bridge University  Press. 

2  See  particularly  A.  L.  Bowley,  Notes  on  Index  Numbers,  Economic  Journal  for  June 
1928,  Sec.  6.     Differences  in  behavior  of  different  classes  of  wages,  both  in  the  short  and 
in  the  long  run,  which  sometimes  amount  to  differences  in  direction,  are,  on  principle,  still 
more  serious  than  the  regional  differences,  which  in  some  cases  are  also  relevant  to  the 
cyclical  process  (migration  of  an  industry  may  be  an  important  innovation;  migration  of 
laborers  from  country  to  towns  and  vice  versa  is  in  part  a  cyclical  phenomenon).     Geo- 
graphically, real  wages  differ  very  much  less  than  money  wages.     The  item  of  rent  is  some- 
times sufficient  to  produce  approximate  equality  in  the  former  over  wide  areas. 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES      567 

Labor's  Share  in  the  national  income  or  in  the  value  of  a  unit  of  product. 
Wage  Bills  and  Rates  must  be  considered  both  in  money  and  in  real 
terms.  For  this  purpose,  they  have  to  be  deflated  by  an  appropriate 
index  of  the  cost  of  living,  an  operation  which  can,  at  best,  yield  but  an 
approximation  to  the  actual  command  over  commodities  enjoyed  by 
the  various  classes  of  workmen  whose  wages  enter  into  the  average. 
There  is,  however,  also  meaning  to  another  concept  of  real  wages — 
which  should  be  called  by  another  name,  say,  Corrected  Wages — viz., 
money  bill  or  rates  divided  by  an  index  of  wholesale  prices.  This 
relation  has,  in  itself,  an  obvious  cyclical  significance.  But  inasmuch 
as  that  index  represents  the  variations  of  the  price  level,  it  acquires  an 
additional  meaning — that  of  an  indicator  of  the  course  of  wages  from 
which  the  influence  of  the  monetary  parameter  has  been  eliminated. 
It  should  be  clear  that  a  cost-of-living  index,  though  in  most  cases  it 
will  give  broadly  similar  results,  is  not  in  principle  qualified  to  render 
that  service  on  account  both  of  the  commodities  that  enter  into  it  and 
of  their  weights. 

The  wage  bill  is  not  the  whole  of  the  income  of  labor,  which  consists 
also  of  other  items  both  in  money  and  in  kind.  But  for  our  problem 
they  do  not  call  for  consideration  in  spite  of  the  fact  that  the  importance 
of  public  expenditure  for  purposes  directly  increasing  the  monetary 
and  real  income  of  the  working  class  increased  considerably  in  the  last 
decades  of  the  epoch  under  survey,  though  not  to  the  extent  that  it  has  in 
postwar  times.  We  similarly  neglect  the  fact  that  wages  do  not  con- 
stitute the  whole  cost  to  firms  of  employing  labor.  Also,  we  assume  that 
the  merely  statistical  effect  on  the  wage  bill  of  small  shopkeepers,  artisans, 
and  so  on,  "migrating"  in  and  out  of  employment  without  changing  the 
economic  nature  of  their  income,  is  negligible.  A  measure  of  variation 
of  real  wage  bill  should  be  made  to  include  increase  in  voluntary  leisure, 
hence,  be  corrected  for  reduction  of  hours  of  work,  provided,  strictly 
speaking,  that  hours  are  not  reduced  in  order  to  distribute  the  burden 
of  unemployment.  Our  failure  to  do  so  involves  understatement  of 
the  historic  rise  in  real  wages.  Average  wage  per  employed  workman 
corrected  for  unemployment  can,  as  long  as  the  population  seeking 
employment  increases,  be  taken  to  indicate  how  the  wage  bill  has  at- 
least  increased  or  at  most  diminished. 

3.  All  wage  series  are  systematic  and  cyclical  and  display  a  descriptive 
trend,  which  is  the  distorted  picture  of  a  result  trend.  What  they 
describe  is  a  primary  and  consequential  phenomenon  that  is  causal 
to  some  of  the  secondary  processes.  The  facts  for  which  we  have  to 
account  are  easily  read  off  from  Charts  XXVIII,  XXIX,  and  XXX. 
It  is  seen  at  a  glance  that  on  the  whole  they  bear  out  the  expectations 
which  our  discussion  of  the  working  of  our  model  would  lead  us  to  form. 


568 


BUSINESS  CYCLES 


The  two  important  exceptions  should  be  disposed  of  at  once.  First, 
we  should  have  expected  money  wages — both  bills  and  rates — not 
indeed  to  fall,  but  to  increase  at  a  smaller  rate  from  1856  to  1873  than 
during  the  preceding  Kondratieff  prosperity.  We  find  in  fact  a  setback 
in  the  late  fifties,  but  a  strong  rise  from  1862  on.  In  America  (see,  also, 
Professor  Mitchell's  weighted  index  of  daily  wages;  but  it  is  safe  to  infer 
that  national  pay  roll  rose  in  at  least  the  same  proportion)  this  is  of 


I860       1865        1870        1875        1880        1885         1890       1895         1900        1905 
CHART  XXVIII.— United  Kingdom  (see  Appendix,  p.  1061). 


1910     I9B 


course  due  to  an  "external  factor,"  the  Civil  War  inflation.  We  have 
discussed  the  point  in  our  historical  survey  and  will  merely  emphasize 
again  that  the  fall  in  the  seventies  did  not  fully  correct  that  deviation. 
Wages  remained  permanently  on  a  new  level,  explainable  by  the  monetary 
event  which  also  influenced  subsequent  cyclical  behavior.  As  far  as 
this  goes,  there  is  close  analogy  with  the  aftermath  of  the  World  War. 
But  we  find  the  same  phenomenon,  though  less  marked,  in  England 
(and  the  rest  of  Europe),  where  it  must  be  attributed  to  gold.  It  is 
significant — both  for  the  theory  and  the  policy  of  wages — that  real  rates 
and  bills  do  not  display  a  corresponding  departure  from  expectation. 
Second,  while  in  this  country  and,  as  far  as  it  is  possible  to  judge,  in 
Germany,  money  bills  and  rates  behaved  according  to  expectation  during 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES      569 


the  prosperity  phase  of  the  third  Kondratieff,  in  England  they  rose 
rather  less  than  we  should  have  expected.  This  tallies  with  other 
peculiarities  of  the  English  situation  of  that  time  but  is  less  easy  to  explain 
than  the  other  case.  It  coincides,  however,  as  has  been  pointed  out 
before,  with  greatly  increased  public  expenditure  and  taxation.  Other- 


1850 


1860  1870  1880  1890  1900 

CHART  XXIX. — United  States  (see  Appendix,  p.  1062). 


1910 


wise,  the  cycles  are  as  we  should  expect.  The  Kitchins  are  not  more 
than  just  recognizable,  the  Juglars  very  well  marked,  the  incomplete 
Kondratieff s  show  in  the  same  way  as  in  clearings.  It  is  particularly 
interesting  to  observe  the  rhythm  in  real  rates  corrected  for  unemploy- 
ment, and  the  way  in  which  Juglars  assert  themselves,  in  the  various 
Kondratieff  phases,  in  the  deviation  from  the  nine-year  moving  average; 
see  Chart  XXXI. 

Bearing  in  mind,  as  we  must  throughout,  that  we  are  dealing  with 
yearly  and,  moreover,  not  very  reliable  figures,  we  observe  that  absolute 
fall  in  the  British  monetary  wage  bill  (Chart  XXVIII)  occurred,  in  the 
two  most  serious  cases,  during  those  years  of  the  depression  phase  of 


570 


BUSINESS  CYCLES 


the  second  Kondratieff  in  which  its  sweep  was  not  interrupted  by  Juglar 
prosperities.  The  figures  for  1874  and  1875  display  but  moderate 
decrease  from  the  peak  of  1873  (485  million  pounds;  1874,  470;  1875, 
465)  but  the  fall  went  on  through  the  Juglar  recovery,  right  to  the 
threshold  of  the  prosperity  of  the  next,  although  it  at  no  time  brought 


1850 


I860 


1870 


1880 


1890 


1900 


1910  1914 


CHART  XXX. — Germany  (see  Appendix,  p.  1062). 


the  wage  bill  back  to  the  figure  of  the  previous  neighborhood  of  equilib- 
rium (365  million  pounds).  This  prolonged  fall  is  not  surprising,  how- 
ever—nor is  the  fact  that  the  next  peak  (1882-1883,  470  million  pounds) 
is  lower  than  the  one  of  1873 — and  perfectly  accords  with  our  cyclical 
schema.  Our  first  impression  to  the  contrary  vanishes  as  soon  as  proper 
account  is  taken  of  the  principle  of  interference,  when  the  phenomenon 
is  readily  recognized  to  be  quite  "regular."  This  is  not  to  say  that  gold 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES      571 

production,  either  directly  or  by  virtue  of  its  responsibility  for  some 
of  the  excesses  of  1872,  had  no  share  in  the  matter,  but  only  that  it 
cannot  have  done  more  than  accentuate  what  would  have  happened 
without  its  decrease  and  what  might  have  been  toned  down  had  it  con- 
tinued to  increase  or  failed  to  increase  previously.  The  figures  for  1884 
and  1885  (the  fall  was  arrested  in  1886,  and  the  1887  wage  bill  increased 
again)  reflect  the  conditions  that  we  may,  according  to  our  schema, 
expect  to  prevail  in  the  depression  of  the  Juglar  at  the  bottom  of  a 
Kondratieff .  But  there  are  within  the  series  four  other  cases  of  absolute 


1850 


I860  1870  1880  1890 

CHART  XXXI.— England  (see  Appendix,  p.  1062). 


fall  in  wage  bill.  The  years  1868  and  189&  do  not  call  for  comment. 
We  also  readily  understand  that  a  fall  may  occur  in  the  course  of  a 
Kondratieff  prosperity,  as  a  result  of  a  "crisis,"  provided  it  be  as  short- 
lived as  it  was  after  1907.  But  there  is  a  prolonged  fall  in  the  English 
figures  for  1901  to  1904,  which  is  contrary  to  expectation  and  to  which 
nothing  corresponds  in  this  country  except  a  small  setback  in  1904. 
It  is  of  course  the  same  phenomenon  that  was  noticed  above,  but  with  a 
new  trait  added.  Most  years  of  unusual  increase  are  easily  identified 
as  years  of  Juglar  prosperities,  though  there  are  some  among  them  that 
belong  to  either  revivals  or  recessions. 

We  do  not  expect,  nor  do  we  find,  any  significant  lag  in  monetary 
wage  bill  behind  comparable  aggregates.  On  the  upgrade  it  rather 
precedes,  if  anything,  the  index  of  wholesale  prices.  There  are  cases, 
such  as  the  behavior  of  the  British  series  in  the  nineties,  that  illustrate 


572  BUSINESS  CYCLES 

almost  ideally,  and  by  so  doing  testify  for,  our  view  of  the  mechanism 
at  work.  In  that  particular  case,  wage  bill  recovers  from  1893  while 
wholesale  prices  are  still  falling  and  breaks  into  a  stronger  rate  of  increase 
in  the  same  year  in  which  wholesale  prices  turn  upward.  In  fact,  as  the 
reader  will  recall  from  our  theoretical  discussion,  the  factors  that  bring 
about  recovery  and  the  different  factors  that  bring  about  prosperity 
may  indeed  make  themselves  felt  at  an  earlier  date  in  some  of  those 
elements  of  the  system  that  move  in  response  to  speculative  anticipation, 
but  their  mechanical  effect  must  show  early  and  unmistakably  in  the 
wage  bill. 

4.  If  there  were  no  unemployment  in  the  neighborhood  of  equilibrium 
from  which  a  cycle  starts,  and  if  population  were  constant,  the  same 
would  be  true  of  money  wage  rates.  Since  these  conditions  are  not  ful- 
filled, we  shall  expect  a  lag  in  rates  on  the  upswing,  which  measures  the 
amount  of  preexisting  unemployment.  We  find  it.  In  1868  wage 
bill  increases  while  rates  remain  constant.  In  1880  wage  bill  goes  up 
while  rates  fall  in  that  and  even  the  following  year.  The  year  1887 
presents  a  similar  picture.  But  we  find,  although  annual  figures  are 
inadequate  for  the  purpose,  rather  less  of  lag  than  we  might  expect.1 
The  English  series  covers  six  Juglar  prosperities,  of  which  four  set  in 
toward  the  end  of  the  year,  so  that  an  annual  average  may  well  produce 
a  spurious  lag,  as  well  as  veil  a  real  one.  In  the  other  two  cases  no  lag  is 
observable,  and  no  lag  greater  than  a  few  months  can  have  been  present. 
The  effect  shows  rather  in  that  wage  bill  rises,  in  the  beginning  of  an 
upswing,  more  strongly  than  wage  rates  and  that  later  on  these  rise 
oftener  than  wage  bill  at  an  increasing  rate.  It  is  worth  mentioning  that 
that  lag,  such  as  it  is,  which  has  been  so  often  emphasized  and  even 
overstated,  docs  not  present  any  new  problem  for  us  and  docs  not  require 
introduction  of  any  cause  extraneous  to  our  mechanism.  But  it  is  still 
more  worth  noticing  that  the  same  holds  true  for  the  "lag"  in  wage 
rates  on  the  downgrade,  which  is  more  pronounced.  Both  bill  and  rate 
increase  in  1866,  1873,  and  1883;  rates  stay  up  in  1884  when  wage  bill 
falls,  and  similarly  in  1893  and  1900.  Wage  rate  does  not  reflect  1907, 
although  wage  bill  does;  neither  reflects  the  recession  of  1913,  in  which 
year  rates  rise. 

Now  most  economists  are  inclined  to  call  upon  such  factors  as  sticki- 
ness in  order  to  explain  this.  Without  denying  the  validity  of  such 
explanations  for  many  cases — although  those  factors  cannot  have  had, 
at  least  up  to  the  turn  of  the  century,  anything  like  the  importance 
they  were  to  acquire  later  on — we  must  still  remember  that,  while  there 

1  The  lag  is,  however,  clearer  in  some  particular  industries.  Compare,  for  example, 
Hooker,  Relation  between  Wages  and  Number  Employed  in  the  Coal  Mining  Industry, 
Journal  of  the  Royal  Statistical  Society,  1894, 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES       573 

is  strong  reason  to  expect  a  fall  in  bills  and  rates  during  "deep"  depres- 
sion, there  is  much  less  reason  for  expecting  that  they  should  fall  in 
recession.  The  opinion  that  wages  should  fall  because,  and  to  the  same 
extent  as,  prices  fall  and  that  their  failure  to  do  so  is  necessarily  an 
additional  source  of  disturbance,  is  incorrect.  If  reality  conformed 
to  our  model,  as  designed  for  the  purposes  of  the  First  Approximation,  the 
monetary  wage  bill  would  indeed  have  to  return  to  its  figure  in  the 
preceding  neighborhood  of  equilibrium,  and  rates  would  consequently 
have  to  fall,  unless  vicarious  unemployment  is  to  ensue,  at  least  pro- 
portionately, but  more  than  that  if  the  population  seeking  employment 
had  meantime  increased.  The  fruits  of  innovation  and  induced  develop- 
ments would  entirely  be  harvested  in  the  form  of  increased  "purchasing 
power"  of  the  unit  of  income.  But  since,  as  we  have  seen,  autodeflation 
never  does  full  work,  and  since  as  far  as  it  does  work,  the  requirements 
of  the  conquest  of  new  economic  space  counteract  its  effects,  there 
need  riot  be  a  fall  in  bill  or  rates  at  all:  we  may  find  instead  a  continued, 
though  weaker,  increase.  Which  we  are  to  expect  in  any  particular 
case,  is,  for  a  normal  recession  and  before  depression  casts  its  shadow, 
indicated  by  the  behavior  of  the  deposit  figure.  If  this  does  not  fall, 
wage  bill  and,  barring  increase  in  population,  rates  need  not.  If  it 
increases,  they  can  and  will1  also  increase,  without  causing  disturbance 
as  they  would  if  failure  to  fall  were  due  to  stickiness  or  increase  were 
due  to,  say,  an  act  of  legislation.  Wages,  bills  or  rates,  need  not  even 
necessarily  fall  during  the  depression  phase,  although  they  are  much 
more  likely  to.  They  will  not,  of  course,  if  the  depression  of  a  cycle 
coincides  with  a  sufficiently  strong  prosperity  or  revival  of  another, 
but  it  is  more  important  to  note  that  they  may  stay  up,  even  barring 
such  coincidence,  if  panics  and  spirals  do  not  play  too  great  a  role. 
In  the  only  Koiidratieff  depression  which  we  have  before  us,  the  English 
wage  bill  actually  increased  by  about  25  per  cent,  and  no  explanation 
external  to  the  mechanism  of  evolution  is  needed  to  account  for  this. 

1  This  is,  of  course,  not  equivalent  to  saying  that  if  the  deposit  figure  be  made  to 
increase,  by  central  bank  action  or  by  any  other  influence  external  to  our  process,  wage  bill 
will  increase  also.  But  we  shall  understand  that  so  pronounced  a  covariation — see,  for 
instance,  the  covariation  of  Pay  Rolls  with  Net  Outside  Deposits  minus  investments  in 
this  country,  Chart  XXIII — should  have  invited  causal  interpretation.  The  fact,  how- 
ever, that  covariation  is  particularly  good  if  deposits  are  cleared  of  the  influence  of  the 
item  most  completely  under  the  control  of  banks — investments — clearly  tells  against  it. 
Perhaps  it  is  not  useless  to  state  once  more  in  a  perfectly  general  form  the  problem  involved. 
We  observe,  let  us  say,  any  time-series  material  consisting  of  the  time  variables  Xi  •  •  •  xn. 
We  find  that  one  of  them  x%  stands  in  an  invariant  relation  to  another,  x,.  It  is  submitted 
that  in  the  absence  of  any  knowledge  about  the  relations  subsisting  between  all  the  variables, 
xi  '  •  •  xnt  we  are  not  justified  in  inferring  from  such  a  finding  that  any  arbitrary  change  in 
Xi  will  be  accompanied  by  the  change  in  xjt  which  actually  does  accompany  it  within  the 
system  observed. 


574  BUSINESS  CYCLES 

What  autonomous  monetary  influences  there  were,  worked  the  other 
way. 

Chart  XXIX  presents  evidence  that,  as  far  as  it  goes  and  as  far  as 
it  can  be  trusted,  justifies  similar  conclusions  for  America.  Railroad 
Earnings  and  Failures  (liabilities)  have  been  inserted  to  provide  links 
with  the  general  business  situation.  The  closeness  of  the  covariation  of 
the  former,  as  well  as  of  "dividends"  (see  Appendix)  with  wage  bill, 
needs  no  emphasis :  the  dominating  cyclical  factor  in  wage  bill  is  of  course 
employment.  Behavior  of  corrected  wage  rates  is  according  to  expecta- 
tion, but  the  strong  increase  in  corrected  wage  bill,  1895  to  1908,  is  not, 
and  must  be  explained  by  peculiarities  of  the  American  environment, 
immigration  among  them. 

5.  A  question,  however,  suggests  itself:  Has  not  the  above  analysis 
gratuitously  assumed  that  the  cyclical  process  leaves  the  relative  marginal 
significance  of  labor  in  the  productive  process  unchanged?  For  only 
in  this  case  it  follows — to  return,  for  the  sake  of  simplicity,  to  the  pure 
model,  also  assuming  population  to  be  constant — that  money  wage 
bill  and  wage  rate  will  be  at  the  same  figures  in  the  new  equilibrium, 
at  which  they  were  in  the  old.  Since  most  innovations  are  not  only 
labor-saving  themselves1  but  induce  adaptations  that  are  also  labor- 
saving,  there  seems  to  be  no  reason  whatsoever  to  expect  this;  hence, 
the  wage  bill  may  decrease  whatever  happens  to  the  total  income.  We 
therefore  meet  again  the  problem  of  machinery  and  labor — it  is  con- 
venient to  confine  ourselves  to  this  special  case — which  we  have  already 
met  twice:  in  our  discussion  of  unemployment  and  in  our  discussion  on 
saving  in  an  otherwise  stationary  process.  But  we  meet  it  now  in  its 
fundamental  aspect,  for  technological  unemployment  is  but  a  special 
and,  moreover,  as  we  have  seen,  a  temporary  form  of  the  effect  of  tech- 
nological improvement  on  the  wage  bill.  We  meet  it,  also,  in  the  new 
setting  provided  by  the  process  of  evolution.  We  cannot  any  longer 
reason  on  invariant  production  functions;  nor  can  we  any  longer  deal 
with  the  isolated  effects  of  laborsaving  innovations.  We  must  take 
account  of  all  the  elements  of  our  process  and,  in  particular,  of  the 
emergence  of  new  investment  opportunities  not  due,  as  it  was  in  the 
case  of  saving,  to  mere  fall  in  the  rates  of  interest.  The  usual  arguments 
both  for  and  against  the  necessity,  likelihood,  or  possibility  that  total 
monetary  wage  bill  will  be  increased  or  decreased  by  technological 
improvement,  thus  become  as  irrelevant  as,  within  their  proper  sphere, 
they  are  inconclusive. 

1  On  the  concept  of  laborsaving  devices,  see  Mr.  Hicks'  Theory  of  Wages,  p.  121. 
Accepting  his  definition,  we  call  an  innovation  laborsaving  if  it  increases  marginal  product 
of  "capital"  more  than  that  of  labor.  It  need  not  actually  diminish  the  latter,  and  cor- 
rected wage  bill  need  not  fall  absolutely. 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES       575 

But  all  the  more  conclusive  are  the  facts.  Although  the  wage  bill 
and  the  share  of  total  income  it  represents  fluctuate  in  the  course  of 
cycles  (even  if  we  exclude  from  total  income,  as  we  should,  gains  and 
losses  from  the  sale  of  capital  assets,  in  particular  from  speculation  in 
land  and  securities),  its  long-run  relation  to  total  income  and  to  most 
other  monetary  aggregates  seems  substantially  stable  (see  Charts 
XXVII  and  XXIII).  We  may  infer  that  an  absolute  amount  of  wage 
bill  would,  as  between  neighborhoods  of  equilibrium,  be  constant, 
if  deposits,  clearings,  and  total  income  were.  This  of  course  implies 
invariance  in  the  long  run,  of  the  money  wage  bill,  both  to  innovation 
as  such  and  to  the  relative  increase,  within  total  output,  of  equipment 
goods  it  induced.  Now  our  data  are  not  too  good  to  make  it  impossible 
to  attack  the  fact.  If,  however,  we  accept  the  evidence,  such  as  it  is, 
then  the  next  possibility  is  to  say  simply  that  it  is  very  remarkable  that 
a  relation  which  theoretically  could  fluctuate  anyhow  yet  remains 
constant,  but  that  it  is  so.  And  this  the  reader  had  better  do,  unless  he 
accept  the  principle  of  the  author's  theory  of  interest,  which  would 
offer  a  theoretical  explanation:  if  it  be  true  that  in  an  ideally  perfect 
competitive  equilibrium  no  assemblage  of  physical  goods  yields  any 
net  return,  then  it  follows  that  there  must  be,  on  the  system's  way  toward 
such  a  state,  even  if  it  be  never  reached,  a  tendency  of  all  incomes, 
except  payments  for  the  services  of  labor  and  natural  agents,  to  vanish. 
The  "wage  and  rent  bill"  would  tend  to  absorb,  with  a  lag,  the  whole  of 
national  income,  and  any  failure  of  wage  bill  to  keep,  in  successive 
neighborhoods  of  equilibrium,  a  constant  proportion  to  it  could  only 
be  due,  either  to  change  in  the  relation  of  marginal  productivity  of  labor 
to  that  of  natural  agents,  or  to  varying  degrees  of  imperfection  in  the 
neighborhoods  for  which  the  figures  of  wage  bill  are  being  compared, 
neither  of  which  would  in  general  be  of  an  importance  sufficiently  great 
to  dominate  the  findings.1  Without  appeal  to  this  theory,  it  is  still 
possible  to  argue  that,  "real  capital"  being  itself  a  product,  its  price 
will  tend  to  be  a  minimum  whenever  the  system  is  approaching  equilib- 
rium. But  if  that  minimum  contains  an  element  of  interest,  no  general 
proposition  is  possible  about  absolute  or  relative  shares  in  the  national 
income,  although  it  remains  true  that  no  such  minimum  principle  operates 
in  the  case  of  wages. 

1  Professor  Leontief  objected  to  the  above  argument  on  the  ground  that,  since  the 
author  does  not  hold  that  interest  or  return  to  real  capital — net  marginal  value  produc- 
tivity— is  actually  zero  in  the  neighborhoods  that  we  observe  in  real  life,  the  result  does  not 
follow,  even  if  his  theory  be  accepted.  This  is  not  so,  however.  If  interest  fails  to  become 
zero  because  it  represents  a  payment  that  is  a  necessary  element  of  the  productive  process, 
a  payment  for  a  distinct  productive  service  for  instance,  then  it  becomes  an  open  question 
what  happens  to  the  relative  importance  of  that  service.  If  interest  fails  to  become  zero 
for  other,  for  instance  frictional,  reasons,  then  there  is  no  room  for  such  an  argument. 


576  BUSINESS  CYCLES 

The  fact,  whether  explained  or  a  mystery,  that  wage  bill  behaves  as 
if  it  were  a  total  monetary  aggregate,  such  as  the  sum  of  household 
incomes,  while  we  know  it  to  be  but  a  variable  part  of  it,  together  with 
what  has  previously  been  said  about  price  levels,  suffices  to  account  for 
the  cyclical  behavior  of  real  and  corrected  wage  bill  and,  with  the  obvious 
qualifications,  of  real  and  corrected  wage  rates.  In  particular,  we  saw 
how  these  quantities  fared  in  recession  and  depression.  That  behavior 
bears  witness  to  the  old  proposition  so  familiar  to  the  common  man, 
that  the  working  class  is,  as  a  whole  and  on  the  whole,  better  off  in 
times  of  falling  than  of  rising  prices.  This  needs  to  be  qualified,  first, 
for  the  fall  of  prices  in  "deep"  depressions;  second,  for  the  rise  of  prices, 
if  any,  in  revivals,  and  third,  for  the  rise  of  prices  in  the  upswing  of  shorter 
cycles  that  lie  in  the  downgrade  of  the  longer  ones.  Nevertheless,  it 
expresses  a  broad  truth.  To  be  sure,  this  truth  is  usually  made  to  rest 
on  mere  stickiness  and  the  simple  arithmetical  effect  of  falling  prices  on  a 
given  income.  This  is  sufficient  only  for  other  than  the  cyclical  cases — 
cases  of  gold  or  government  inflation,  for  instance — which  are  historically 
associated  with  misery.  But  our  argument  shows  that  there  is  some 
foundation  for  it  also  with  reference  to  cyclical  variations  in  price  level. 
As  in  similar  cases,  the  true  contours  can  be  expected  to  reveal  them- 
selves most  clearly  in  the  Kondratieff.  So  they  do.  Information  is 
available  for  England,  that  would  not  warrant  positive  assertion,  yet 
suggests  that  this  holds  beyond  the  span  covered  by  our  series.  Real 
wage  bill,  as  well  as  rates,  very  probably  rose  somewhat  from  1775  to 
1815,  that  is  to  say,  in  the  later  stages  of  a  Kondratieff  revival,  the  whole 
of  a  Kondratieff  prosperity,  and  in  a  little  more  than  a  Kondratieff 
recession  that  was  distorted  by  government  inflation  and  hence  cannot 
be  accepted  as  normal.  Thereafter  there  was,  until  the  forties,  little 
change  in  incomes,  while  cost  of  living  fell  so  considerably  that  it  is  not 
likely  that  better  information  about  unemployment  would  invalidate 
the  inference  that  real  wage  bill,  as  well  as  rates,  rose  much  more  than 
they  had  risen  before.  In  the  forties  real  rates  cannot  have  increased 
considerably.  During  the  period  of  rising  prices  that  followed,  com- 
pleting the  Kondratieff  upgrade,  but  outlasting  it  for  reasons  we  know, 
corrected  rates  certainly,  and  real  rates  possibly,  fell  (see  Chart  XXVIII) 
until  the  top  of  the  grade  was  reached — though,  because  of  increased 
employment,  the  real  wage  bill  almost  certainly  rose — and  then,  from 
the  top  (1856)  to  1897,  took  the  stride — nearly  80  per  cent — which 
created  the  modern  standard  of  life.  There  was  hardly  any  increase 
in  real  rates  in  the  period  1898  to  1913  taken  as  a  whole,  though  of  course 
there  was  in  real  wage  bill.  Normally,  and  with  but  a  moderate  number 
of  exceptions,  real  bill  and  rates  increase  in  season  and  out  of  season.  But 
it  is  the  increase  in  the  three  last  phases  of  each  of  our  two  Kondratieffs 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES       577 

that  is  mainly  responsible  for  the  observed  trend  in  real  rates.  Since 
this  trend  is  the  effect,  however  distorted  by  external  factors,  of  the 
working  of  our  mechanism,  it  may  also  be  identified  as  a  result 
trend. 

From  the  above  analysis  it  may  be  inferred  that  the  sequence  of 
cyclical  situations  tends  to  produce  wage  rates  and  wage  bills  which  are, 
with  due  qualifications  as  to  the  many  cases  of  indeterminateness  that 
arise,  adapted  to  them  in  the  sense  that  they  do  not  in  turn  require  or 
induce  a  distinct  process  of  adaptation  on  the  part  of  other  elements  of 
the  system.  In  particular,  it  follows  that  occurrence  of  the  upper  turning 
point  or  of  depression  is  not  normally  such  an  adaptation  to  preceding 
wage  rates,  enforced,  for  example,  by  any  failure  of  the  "purchasing 
power"  of  laborers  to  develop  in  a  manner  that  would  enable  them  to 
buy  the  increment  of  product  resulting  from  every  step  in  the  evolu- 
tionary process.  Nor  is  it  true  that  money  wage  rates  or  bills  would 
have  first  to  rise,  or  to  be  increased  by  public  authority  or  by  the  pressure 
of  organized  labor  before  the  system  can  move  up  from  the  lower  turning 
point.  On  the  other  hand,  we  have  also  seen  that  the  mere  fact  of  wages 
rising  in  prosperity  to  the  extent  appropriate  to  the  upward  shift  of 
"demand  curves  for  labor"  is  not  the  "cause"  that  turns  prosperity  into 
recession,  and  that  the  failure  of  money  wage  rates  to  fall  during  the 
latter  phase  has  normally  as  little,  if  anything,  to  do  with  pushing  the 
system  into  depression  as  such  fall  as  occurs  in  depression  has  to  do  with 
helping  the  business  organism  on  to  the  path  of  recovery. 

These  statements,  however,  do  not  imply  anything  about  the  effects 
of  those  levels  and  variations  of  money  wage  rates  that  are  not,  as  we 
have  put  it,  produced  by  the  working  of  our  process  but  are  imposed  on 
the  system.  Such  imposed  rates  may — although  they  need  not;  there 
are  plenty  of  reasons  why,  in  a  given  instance,  the  system  should  fail  to 
give  effect  to  its  tendency  to  produce  the  adapted  rate— spell  disturbance, 
i.e.,  enforce  additional  adaptations  of  other  elements  of  the  system. 
But  even  if  they  do,  this  in  itself  only  means  that  they  disturb  a  process 
which  by  virtue  of  its  very  nature  creates,  although  it  also  eliminates, 
disequilibria.  Interference  with  such  a  process  need  not  intensify  these 
disequilibria  but,  on  the  contrary,  may  mitigate  them.  We  have  not 
even,  by  denying  that  the  cyclical  behavior  of  wages,  such  as  our  process 
tends  to  produce,  is  causal  to  the  occurrence  of  recessions,  depressions, 
and  revivals,  completely  prejudged  the  question  whether  the  first  two 
could  not  be  toned  down  and  the  third  facilitated  by  the  imposition  of 
appropriate  "inadapted"  wage  rates.  No  particular  dignity  is  claimed 
for  the  adapted  rates.  Our  analysis  refutes,  indeed,  some  of  the  argu- 
ments that  are  commonly  adduced  in  favor  of  high  or  low  wage  policies. 
But  it  does  not  preclude  others.  Satisfactory  treatment  of  this  problem 


578  BUSINESS  CYCLES 

requires,  however,  the  whole  apparatus  of  general  theory  and  cannot 
be  attempted  here.  And  only  a  few  aspects  of  it  will  be  presented  in  our 
discussion  of  the  postwar  epoch. 

D.  Deposits  and  Loans. — Returning  to  the  argument  of  Sec.  B,  we 
will  now  discuss  the  behavior  of  the  sources  of  system  expenditure,  i.e., 
of  (the  cash  and)  the  balances  or  "deposits"  of  firms  and  households, 
and  thereby  also  gather  up  the  threads  of  the  analysis  of  Chap.  Ill, 
Sec.  D,  as  developed  in  Chap.  IV.  Perhaps  it  is  not  superfluous,  speaking 
of  " sources"  of  expenditure,  to  advert  to  the  metaphorical  character  of 
that  phrase  and  to  the  misleading  associations  to  which  it  may  give  rise. 
We  are  not  now  moving  toward  the  origin  of  the  cyclical  process  but,  on 
the  contrary,  away  from  it.  It  is  precisely  in  order  to  avoid  any  implica- 
tions about  mechanical  effects  being  exerted  on  the  pulse  of  business  by 
the  "flow  of  funds"  that  the  subject  of  balances  has  not  been  taken  up 
before  the  subject  of  the  expenditure  they  finance.  For  the  same  reason, 
discussion  of  the  behavior  of  banks  and  of  the  role  of  central  banks  will 
be  deferred  to  a  still  later  stage.  In  this  section  we  are  going  to  look 
at  financing  from  the  standpoint  of  firms  and  households  only. 

1.  Any  act  of  expenditure  may  be  financed: 

a.  By  Previous  Receipts.  The  reader  will  recall  the  proposition  that 
in  a  process  which  merely  reproduces  itself  at  constant  rates  (stationary 
process)  all  acts  of  expenditure  could  and  would  be  so  financed.  Now 
the  bulk  of  a  nation's  business,  at  any  given  point  of  time,  consists  of 
transactions  that  repeat  with  adaptive  variations  the  transactions  of 
previous  periods  of  account.  For  some  purposes  it  is  convenient  to  split 
each  of  these  transactions  into  an  exact  repetition  and  a  plus  or  minus 
variation,  and  to  look  upon  all  these  exact  repetitions,  on  the  one  hand, 
and  upon  all  the  corresponding  variations,  on  the  other  hand,  as  separate 
classes  of  transactions  to  be  distinguished  from  a  third  class,  which  con- 
sists of  new  transactions — mainly,  those  effected  by  entrepreneurs  or 
induced  by  entrepreneurial  activity.  At  the  level  of  abstraction  appro- 
priate to  our  pure  model  we  applied  that  proposition  to  the  first  class 
of  transactions,  the  repetitions.  But  we  cannot  do  so  any  longer.  We 
have,  indeed,  to  recognize  the  fact,  much  more  frequently  observable  in 
Continental  Europe  than  in  England  or  the  United  States,  that  many  old 
firms  actually  run  their  business  on  "owned"  cash  or  "owned"  deposits; 
but  this  is  now,  even  for  exact  repetitions,  merely  the  limiting  and  not 
the  general  case:  most  firms,  old  as  well  as  new,  currently  borrow  and 
repay,  even  within  the  most  ordinary  business  routine. 

This  practice  is,  as  we  have  seen  (Chap.  Ill),  a  by-product  of  the 
process  of  evolution  through  which  borrowing  and  credit  creation  intrudes 
into  the  old  strata  of  the  economic  system,  partly  because  balances 
created  for  the  purpose  of  financing  innovation  are  hardly  ever  completely 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES      579 

eliminated  by  autodeflation,  partly  because  other  balances  are  created 
for  the  purpose  of  financing  the  expansion  of  the  noninnovating  sectors 
which  comes  about  in  response  to  the  impulse  given  by  innovation.  What 
we  have  called  the  Conquest  of  New  Economic  Space  thus  expands 
permanently,  though  not  in  proportion  to  physical  output,  the  monetary 
system  and  at  the  same  time  creates  that  gap  between  producers'  expendi- 
ture and  receipts  which  it  then  becomes  the  most  "regular"  business  of 
banks  to  bridge.  A  little  reflection  will  show  that  this  gap,  which  is  so 
familiar  a  phenomenon  as  to  seem  hardly  to  stand  in  need  of  explanation, 
is  in  fact  entirely  due  to  "progress"  and  would  not,  except  as  a  result  of 
previous  waves  of  progress,  be  present  in  a  stationary  society.  But, 
although  not  extraneous  to  our  model,  the  fact  that,  in  a  society  the 
monetary  process  of  which  has  become  adapted  to  evolution,  repetitions 
may  require  financing  as  much  as  plus-variations  or  new  transactions 
affects  our  expectation  as  to  the  behavior  of  balances.  So  far  as  repeti- 
tions are  actually  financed  by  owned  cash  or  original  deposits,  this  cash 
or  these  deposits  do  not,  in  consequence  of  a  decision  to  restrict  opera- 
tions, disappear  but  simply  become  idle.  But  so  far  as  they  are  financed 
by  borrowing,  the  corresponding  balances  do  not  become  idle  but  dis- 
appear in  that  case  through  repayment  or  else  fail  to  emerge  altogether. 
The  businessman's  decision  and  its  economic  results  are  the  same, 
whether  it  is  "quantity"  or  "velocity"  of  deposits  that  is  affected,  but 
the  statistical  picture  is  different. 

While,  however,  we  have  to  recognize  that  mere  repetitions  also 
present  a  distinct  problem  of  financing,  the  problem  of  financing  new 
transactions  may  frequently  be  solved  by  previous  receipts  being  deflected 
from  the  channel  into  which  they  have  previously  been  directed.  A 
man,  who  has  been  in  the  habit  of  setting  aside  a  yearly  sum  in  order  to 
buy  a  new  motorcar  every  five  years,  may  at  the  end  of  one  of  these 
five-year  periods  decide  to  buy  an  aeroplane  instead.  Not  only  system 
but  also  consumers'  expenditure  or,  had  we  chosen  an  example  from  the 
sphere  of  production,  producers'  expenditure  remains  unaffected  in  case 
of  such  Deflection. 

6.  By  Overspending.  This  term  is  to  mean  allowing  one's  balance  to 
fall  below  the  amount  appropriate  to  the  requirements  in  the  previous 
neighborhood  of  equilibrium.  This  will  affect  system  expenditure  but 
not  total  balances.  The  opposite  case,  Underspending  or,  as  we  have 
previously  called  it,  Nonspending,  occurs  if  cash  balances  rise  above 
that  amount  and  is  usually  referred  to  as  "Hoarding."  This  word, 
however,  lends  a  wrong  color  to  an  extremely  simple  state  of  things  and 
should  never  be  used  except  in  order  to  designate  the  thing  it  really 
means — still  observable  in  India,  for  instance — which,  sociologically  and 
economically,  is  perfectly  definite  and  presupposes  a  characteristic 


580  BUSINESS  CYCLES 

attitude  toward  money  entirely  foreign1  to  capitalist  society.  We  also 
recall  that  it  is  misleading,  although  for  other  reasons,  to  express  the  fact 
of  underspending  by  reference  to  any  demand  for  balances,  which  is  in 
this  case  purely  fictitious,  or  to  call  it  saving.  Release  from  under- 
spending  is,  of  course,  an  important  method  of  financing  the  expansion  of 
recovery.  It  should  be  observed  that,  although,  given  the  sum  total  of 
cash  items  of  households  and  firms,  all  of  them  cannot  increase  or  decrease 
at  the  same  time,  yet  the  inactive  part  of  them  can. 

c.  By  Selling  Assets.     If  an  asset  be  sold  to  a  bank  (to  this  we  refer 
as  "member  bank  investment")*  customers'  balances  will  be  increased 
unless  the  bank  neutralizes  the  effect.     And  if  it  be  done  in  order  to 
finance  an  act  of  expenditure,  it  will  of  course  lead  to  an  increase  in 
system  expenditure,  although  it  would  not  by  itself  and  in  the  absence 
of  such  intention  have  that  effect.     The  answer  to  the  question  by  how 
much  it  will  increase  it  depends  on  the  behavior  of  those  firms  and 
households  that  are  the  recipients  of  the  sum  thus  procured — hence 
on  the  cyclical  phases  into  which  that  event  happens  to  fall.     If  an  asset 
be  sold  to  another  firm  or  household,  it  will  not  increase  the  sum  of 
balances  but  may  still  increase  system  expenditure  if  it  is  the  occasion 
of  turning  an  inactive  deposit  into  an  active  one.     If  a  bank  sells  to 
firms  or  households,  the  balances  of  the  public  are  decreased. 

d.  By  "  Temporary  Investment"     This  is  the  term  given  in  Chap.  Ill, 
Sec.  A,  to  signify  using,  or  borrowing  from  another  firm  or  household,  a 
sum  which  is  and  remains  intended  for  another  act  of  expenditure  lying 
sufficiently  far  off  to  make  it  possible  to  replace  or  repay  that  sum  before  it 
will  be  wanted.     We  assume  in  sufficient  approximation  to  fact  that  Tem- 
porary Investment  acts  exclusively  through  the  open  market  or  the  stock 
exchange.     It  increases  system  expenditure  but  not  the  total  of  balances. 

e.  By  Issuing  Promises  to  Pay.     These  may  be  issued,  for  example, 
in  the  form  of  bills  of  exchange,  provided  they  are  actually  accepted  in 
payment  and  circulate  (see  Chap.  Ill,  Sec.  D,  3).     This  item  we  shall 
neglect,  however. 

/.  By  mining  or  importing  the  monetary  metal  and  having  it  coined,  or 
exchanging  it  for  a  deposit,  or,  what  economically  comes  to  the  same  thing 
(as  has  been  pointed  out  by  Wicksell),  forging  money.  This  increases 
deposits  and  the  bank's  cash  by  the  same  amount.  If  not  deposited, 
the  money  may  only  increase  the  outside  circulation. 

g.  By  Borrowing  from  a  Bank.  This,  in  England  and  the  United 
States,  increases  deposits  immediately  (with  a  qualification  as  to  over- 

1  The  only  cases  in  capitalist  society  at  all  akin  to  true  hoarding  are  those  in  which 
lawful  money  is  held  from  a  distrust  of  banks,  or  some  kind  of  lawful  money  from  distrust 
of  other  kinds,  or  some  commodity  from  distrust  of  all  kinds.  But  these  phenomena  are  in 
a  class  by  themselves  and  entirely  different  from  nonspending. 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES      581 

drafts),  and  in  Germany  as  soon  as  payments  to  firms  or  households  are 
effected. 

h.  By  using  one's  own,  or  borrowing  some  other  household's  or  firm's 
uninvested  savings  or  accumulations.  By  borrowing  we  also  mean  to 
include  methods  which  do  not  technically  come  within  that  term,  such 
as  the  formation  of  partnerships. 

1.  By  Taxation  and  the  Issue  of  Government  Fiat.     As  a  method  of 
financing  business  these  items  will  now  be  neglected,  although  we  have 
repeatedly  met  them  in  this  role  in  the  historical  chapters. 

j.  By  Buying  on  Credit.  But  we  will  in  general  assume,  although 
this  is  often  far  from  the  truth,  that  in  this  case  sellers  borrow  from  banks 
equivalent  amounts.  The  same  applies  to  the  opposite  case,  in  which 
the  burden  of  financing  production  is  shifted  by  payment  being  exacted 
in  advance  of  delivery. 

k.  By  Borrowing  from  Abroad.  This  acts  exactly  like  /  if  it  leads  to 
imports  of  gold,  and  in  exactly  the  opposite  way  if  the  proceeds  of  the 
loan  buy  foreign  commodities. 

2.  Since  all  these  methods  of  financing  combine  to  shape  the  behavior 
of  balances  in  the  phases  of  the  cyclical  process,  and  since  there  are  so 
many  external  factors  that  are  bound  to  assert  themselves,  it  becomes 
doubtful  whether  the  contours  to  be  expected  from  our  model  will  show 
at  all  in  the  series  of  deposits  and  of  bank  loans.     These  expectations 
need  not  be  restated.     Firms'  and  household's  (cash  plus)  balances  are 
primary  and  consequential  and  make  a  natural,  systematic,  and  cyclical 
series,  which  should  behave  like  system  expenditure  and  display  a  result 
trend  in  the  same  sense.     The  same  holds  for  member  banks'  loans  and 
discounts.     Modifications  are  necessary  even  in  the  case  of  the  pure 
model  and  will  be  noticed  presently.     But  it  is  interesting  to  see  how  far 
reality  conforms  to  what  seems  to  be  a  theoretical  construction  very 
remote    from    actual    fact.     Chart    XXV    supplies    the    answer.     The 
covariation  of  deposits,  loans,  clearings,  and  of  all  three  of  them  with  pig- 
iron  consumption  and  the  production  of  equipment,  though  far  from 
perfect,  is  striking.     The  reader  is  invited  to  satisfy  himself  that  it  could 
hardly  be  expected  to  be  stronger,  the  material  being  what  it  is,  if  there 
were  no  factors  or  mechanisms  at  work  other  than  those  embodied  in 
our  model.     Our  problem  is  not  so  much  how  to  explain  deviations  from 
that  expectation  as  how  to  explain  a  conformity  so  much  beyond  reason- 
able hope. 

For  both  these  purposes  it  is  necessary  to  bear  in  mind  the  short- 
comings of  our  material,  which  make  it  impossible  to  place  much  con- 
fidence in  anything  except  the  roughest  contour  lines.  Our  analysis  is 
based  mainly  on  American  data,  neither  English  nor  German  statistics 
being  at  all  adequate  for  our  purpose,  although  they  may  be  used  as 


582  BUSINESS  CYCLES 

supporting  evidence  and  although  more  valuable  information  is  available 
for  Germany  since  1901.  *  Those  American  data  refer  to  national  banks2 
and  thus  represent  a  sample  of  varying  importance,  which  cannot  be 
trusted  to  reflect  the  condition  of  all  commercial  banks.3  The  series  of 
individual  deposits  less  clearing-house  exchanges,  which  is  the  one  that 
has  been  primarily  used,  substantially  gives  the  balances  of  firms  and 
households  that  we  want,  since  it  excludes  government  and  interbank 
deposits — although  deposits  due  to  savings  banks  were  included  by 
some  national  banks  until  Apr.  26,  1900 — but  the  device  of  using  individ- 
ual deposits  of  national  banks  outside  New  York  in  order  to  reduce 
the  influence  of  speculative  and  other  transactions  which  we  wish  to 
exclude  is  of  course  as  unsatisfactory  in  this  case  as  it  was  in  the  case  of 
clearings.  The  fact  that  short  fluctuations  in  deposits  and  in  loans  and 
discounts  outside  New  York  City  are  inversely  related  to  those  in  New 
York  deposits  and  loans  and  discounts  tends,  however,  not  only  to 
reassure  us  on  that  point,  but  also  to  show  that  the  bankers'  bank  func- 
tions, which  we  attribute  to  the  national  banks  of  New  York,  was  not 
completely  overshadowed  by  their  commercial  and  industrial  business. 
In  comparing  this  series  of  outside  deposits  with  our  clearing  series, 
which  represents  a  different  sample,  we  are,  of  course,  opening  up  another 
source  of  error. 

1  See  Professor  A.  Hahn's  analysis  in  Vierteljahrshefte  zur  Konjunkturforschung,  1927. 
German  bookkeeping  methods  and  German  terminology  increase  the  difficulty  of  inter- 
preting German  banking  statistics,  which  for  the  nineteenth  century  are  also  unsatisfactory 
for  other  reasons.     What  the  writer  believes  to  be  acceptable  indicators  of  the  behavior  of 
balances  plus  hand-to-hand  circulation  are,  however,  plotted  on  the  pulse  charts. 

2  National  bank  data  have  been  analyzed  by  many  students.     There  is  however  one 
outstanding  contribution,  to  which  the  writer  wishes  to  acknowledge  obligation  and  to 
refer  the  reader  for  many  points,  regional  differences  among  them,  that  cannot  be  dealt 
with  in  the  text:  A.  A.  Young,  An  Analysis  of  Bank  Statistics  for  the  United  States,  reprint 
of  articles  from  the  Review  of  Economic  Statistics,  1928,  especially  pp.  1-7,  21-32.     Also 
see  J.  P.  Norton,  Studies  in  the  New  York  Money  Market,  1907,  and  O.  M.  Sprague,  Crises 
under  the  National  Banking  System.     There  are,  of  course,  figures  for  other  banks,  both 
since  and  anterior  to  the  establishment  of  the  National  Banking  System.     Some  of  the 
older  material  has  been  generously  put  at  the  writer's  disposal  by  Professor  Cole.     Refer- 
ence should  be  made,  also,  to  W.  B.  Smith  and  A.  H.  Cole,  op.  cit.     Estimates  of  total, 
time,  and  demand  deposits,  1834-1935,  have  been  made  by  Dr.  C.  E.  Thomas;  they  will  be 
presented  in  the  writer's  book  on  money.     For  the  purpose  in  hand  it  has  seemed  best  to 
keep  to  comparatively  safe  ground.     No  indications  have  been  encountered  to  warrant 
a  suspicion  that  results  are  affected  thereby. 

8  The  growing  importance  of  trust  companies  in  the  nineties  and  particularly  after 
1899,  which,  with  the  rise  of  the  Kondratieff,  "began  to  invade,  on  an  extensive  scale,  the 
field  of  deposit  banking"  (A.  D.  Noyes,  Forty  Years  of  American  Finance,  1909,  p.  367) 
and  increased  their  deposits  from  $198,000,000  in  1898  to  $834,000,000  in  1906  (ibid., 
p.  368),  constitutes  a  further  and  very  serious  limitation  of  the  symptomatic  value  of 
national  bank  figures. 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES       583 

Moreover,  for  our  purpose  we  ought  to  include  overdrafts  with 
deposits,  because  from  the  standpoint  of  the  individual  firm  or  household 
they  are  no  less  available  "funds"  than  deposits.  They  fluctuated, 
however,  as  deposits,1  only  much  more  strongly,  and  hence  but  accentuate 
the  phenomena  under  discussion.  But  they  are  included  in  loans  and 
discounts  prior  to  Dec.  1,  1898.  Collateral  loans  can  be  separated  from 
others  since  1892.  This  has  not  been  done,  because,  as  mentioned  in 
the  first  chapter,  they  may  and  often  do  serve  purposes  other  than  stock 
exchange  speculation.  Concerning  time  deposits,  reasons  will  be  offered 
later  in  this  section  and  in  our  discussion  of  postwar  developments  for 
believing  that  our  inability  to  separate  them  from  demand  deposits2  is 
less  unfortunate  than  one  would  think.  For  the  period  at  present  under 
discussion  time  deposits  in  national  (and  all  commercial)  banks  were  not 
very  important  until  1896.  After  that  date  they  rose  rapidly,  and, 
although  in  1915  they  stood  at  a  level  which  was  low  when  compared  with 
the  figures  to  which  they  were  to  soar  in  the  postwar  period,  they  cannot 
be  considered  as  negligible.  As  a  matter  of  banking  practice,  however, 
it  seems  plausible  that  during  the  prosperity  of  the  late  nineties,  which 
they  had  to  face  in  fetters  that  were  comparatively  tight,  national  banks 
adopted  a  policy  of  attracting  along  with  genuine  savings  also  nonsaving 
accounts  by  offering  to  treat  them  like  saving  deposits.  If  that  was  so, 
we  probably  lose  little  by  using  total  deposit  figures,  without  attempting 
to  correct  them  for  time  deposits. 

Since,  finally,  cash  in  hand  and  balance  with  a  bank  are,  normally 
and  for  the  individual  firm  or  household,  but  different  forms  of  the  same 
thing,  the  really  significant  item  is  cash  plus  balances  rather  than  balances 
alone.  Substantially,  however,  we  must,  at  least  for  prewar  times, 
content  ourselves  with  the  latter.  The  only  indication  we  have  about 
cash  in  hand  is  from  Money  in  Circulation,  which,  besides  lacking  com- 
parability with  national  bank  deposits  outside  New  York  City,  must  be 
estimated  in  extremely  hazardous  ways.  We  have  a  continuous  but 
highly  unreliable  series  of  money  in  the  country  outside  the  Treasury 
(specie,  bank  notes,  treasury  notes,  and  so  on)  from  1813  to  date,3 

1  Cf.  L.  W.  Hall,  Cycles  in  Banking,  p.  74. 

2  That  inability  is  not  absolute.     Professor  Mitchell  (Business  Cycles,  1918,  p.  821) 
has  presented  an  estimate  of  time  deposits  back  to  1890.     Dr.  Thomas's  estimate  comprises 
deposits  evidenced  by  savings  passbooks,  time  certificates  of  deposits,  and  postal  savings 
deposits  back  to  1834,  and  thus  implies,  to  be  sure,  a  definition  of  time  deposits  that  differs 
from  the  usual  and  official  one.     But  for  1931  these  three  categories  represented  about  90 
per  cent  of  the  total  time  deposits  reported  by  the  Comptroller.     And  they  probably  over- 
state rather  than  understate  that  amount,*  which  really  differed  in  nature  from  demand 
deposits.     For  early  data  relating  to  Mutual  Savings  Banks  see  E.  W  Keyes,  A  History  of 
Mutual  Savings  Banks  in  the  United  States,  1876. 

8  Cf.  A.  Barton  Hepburn,  History  of  Currency  in  the  United  States,  1915,  for  early 


584  BUSINESS  CYCLES 

from  which  the  series  of  money  in  circulation1  has  to  be  derived  by 
subtracting  money  in  banks.  For  the  period  from  1834  to  1863  the 
Comptroller  of  the  Currency  recorded  both  items.  After  1900,  when 
about  90  per  cent  of  the  state  banks  and  trust  companies,  though  only 
from  20  to  25  per  cent  of  private  banks,  reported  either  to  state  authori- 
ties or  directly  to  the  Comptroller,  we  have  figures  which  cannot  be  far 
from  the  truth  and  are  amenable  to  fairly  convincing  corrections.  But 
before  1834  and  during  the  troubled  period  from  1864  to  1875  estimation 
becomes,  to  say  the  least,  very  risky.  For  1875  to  1900  an  intermediate 
situation  prevails.  The  series  of  money  held  by  reporting  banks  is  more 
significant  after  than  before  1875.  But  for  an  estimate  of  money  held 
by  nonreporting  banks  we  are  lacking  the  indication  given  since  1900 
by  the  Treasury's  estimate  of  deposits  in  all  banks.  There  are  various 
estimates,  among  which  Professor  Mitchell's — from  the  average  deposits 
of  reporting  private  banks,  1890  to  1911 — seems  to  hold  first  place. 
We  might  also  extrapolate  backward  the  ratio  between  deposits  in  all 
and  in  reporting  banks  for  1900  to  1914  and  apply  the  result  to  that 
section  of  the  series  of  money  in  reporting  banks.  In  any  case,  however, 
it  is  a  matter  of  estimating  from  estimates  by  means  of  assumptions, 
some  of  which  are  obviously  contrary  to  fact. 

3.  If,  in  order  to  start  from  a  reasonably  simple  model,  we  now  focus 
attention  on  items  a,  b,  c,  g,  and  h  of  our  list,  we  have  first  to  qualify 
our  expectation  as  to  perfect  parallelism  between  the  variations  of" 
outside  clearings  and  outside  individual  deposits  by  taking  account  of 
the  phenomena  noticed  under  a  and  b.  In  fact,  if  there  are  in  the  system 
any  deposits  at  all  which  do  not  emerge  through  loans  and,  hence,  do 
not  disappear  through  repayment,2  expectation  as  to  volume  of  deposits 
must,  even  for  the  pure  model,  be  supplemented  by  an  expectation  as  to 

data,  which  can  be  supplemented  by  the  annual  reports  published  by  the  Comptroller  of 
the  Currency;  see,  in  particular,  those  of  1900,  1904,  1920,  and  1931.  The  quality  of  the 
series  differs,  of  course,  as  between  different  periods.  But  it  never  reaches  reliability,  the 
components  of  its  items,  except  annual  coinage,  being  rough  estimates.  For  the  correction 
made  in  1907  see  Annual  Report  of  the  Director  of  the  Mint,  1907. 

1  See  estimates  by  Professors  Kemmerer,  1897-1904;  Irving  Fisher,  1896-1909;  Wesley 
C.  Mitchell,  1890-1911;  W.  I.  King,  1880-1920;  A.  A.  Young,  1901-1914.     For  all  these, 
as  well  as  for  another  estimate,  1900-1926,  cf.  Y.  S.  Leong,  An  Estimate  of  the  Amount  of 
Money  Held  by  the  Banks  and  of  the  Amount  of  Money  in  General  Circulation,  Jounal 
of  Political  Economy,  1927.     The  writer  acknowledges  the  help  derived  from  a  report  by 
Dr.  C.  E.  Thomas,  who  made  the  attempt  to  estimate  money  in  circulation  for  1813  to  1933. 
For  our  purpose  it  was  not  thought  necessary  to  reproduce  the  chart. 

2  Balances  that  arise  from  loans  need  not  always  be  repaid  when  they  are  no  longer 
wanted.     And  balances  which  result  from  customers'  selling  assets  to  banks  may,  although 
" owned,"  disappear  in  that  case,  because  their  owners  may  prefer  to  buy  assets  instead  of 
keeping  them  idle.     The  statement  in  the  text  assumes  that  the  possibilities  of  borrowed 
deposits  becoming  idle  and  of  owned  deposits  being  eliminated  are  of  minor  importance 
for  the  argument  in  hand. 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES       585 

the  rate  of  spending.  Obviously,  there  will  be  Overspending  in  pros- 
perity, return  to  normal  amounts  of  (cash  plus)  balances  in  recession, 
Underspending  in  depression,  and  again  return  to  normal  habits  in 
recovery.  Clearings  should  therefore  rise  more  in  prosperity  and  fall 
more  in  depression  than  balances  do.  Any  increase  in  recession  should 
be  more  marked,  the  increase  in  recovery  less  marked,  in  the  deposit 
than  in  the  clearing  series.  As  far  as  our  Chart  (XXV)  can  be  trusted  at 
all,  its  testimony  may  on  the  whole,  due  attention  being  paid  to  inter- 
ference of  cycles  with  each  other,  be  invoked  in  support.  It  would 
be  hazardous  to  go  further  and  to  interpret  what  at  first  sight  gives 
the  impression  of  a  long-time  tendency  of  the  curves  to  draw  apart, 
in  the  light  of  cyclical  effects  only.  Too  many  other  elements  enter  into 
it.  It  is,  however,  permissible  to  point  out  that,  as  far  as  that  impression 
is  due  to  the  behavior  of  the  curves  in  the  last  years  of  our  period,  it  is 
in  fact  accounted  for,  in  part  at  least,  first  by  a  Juglar  and  then  by  the 
beginning  of  a  Kondratieff  recession. 

We  might  discard  the  problems  incident  to  the  growing  divergence  of 
clearings  and  deposits  by  eliminating  the  "trends"  in  both  series — two 
straight-line  trends  fitted  to  each  of  them  for  the  intervals  which  belong 
to  the  two  Kondratieffs  would  serve  the  purpose — and  then  derive  a 
strongly  cyclical  series  of  deviations  of  clearings  divided  by  deviations  of 
deposits,  which,  although  open  to  criticism  on  the  score  of  the  imperfect 
comparability  of  the  components,  as  well  as  to  other  objections,  could 
be  said  to  indicate  roughly  the  variations  of  the  rate  of  spending  in  the 
shorter  cycles.  It  might  serve  to  interpret  certain  properties  of  the 
relation  between  deposits  and  price  level  (see  the  Pulse  Charts  V,  VI, 
VII).  Recalling  the  results  we  finally  arrived  at  about  the  cyclical 
behavior  of  output,  we  should,  irrespective  of  overspending  and  specula- 
tive anticipation,  expect  the  price  level  to  rise  somewhat  less  than  deposits 
in  prosperity.  Because  of  overspending  it  should,  irrespective  of  such 
increase  of  output  as  actually  occurs,  rise  somewhat  more.  There  is  no 
reason  to  believe  that  these  two  effects  will  exactly  balance  so  as  to  make 
price  level  vary  in  proportion  to  the  variation  of  deposits;  but  they  will 
balance  to  some  extent.  In  deep  depression  again,  underspending, 
intensifying  the  effects  of  decrease  of  deposits,  is  partly  at  least  counter- 
acted by  decrease  in  output.  The  rates  of  spending  and  of  output  also 
vary  in  the  same  sense  in  revival,  and  only  in  recession  do  they  fail  to  move 
together.  Since,  however,  the  latter  effect  is  partly  suppressed  by  trend 
elimination  and,  in  any  case,  weak  within  the  shortest  cycles,  we  shall 
not  share  Professor  Pigou's  astonishment1  at  Mr.  Carl  Snyder's  finding 
that  "deposit  velocity"  and  "trade  activity"  display  a  tendency  to 

1  "  Industrial  Fluctuations,"  Chap.  XV.  On  his  analysis  of  English  figures  see  Edie 
and  Weaver,  Velocity  of  Bank  Deposits  in  England,  Journal  of  Political  Economy  1930, 
p.  898.  Mr.  Keynes  discussed  Mr.  Snyder's  finding  in  Treatise  on  Money,  pp.  80-82. 


586  BUSINESS  CYCLES 

neutralize  each  other's  short-time  effects  on  the  price  level,  although  we 
may  be  unable  to  follow  Mr.  Snyder  when  he  concludes  that  "neither 
variations  in  velocity  nor  in  trade  activity  are  normally  a  factor  in  the 
determination  of  the  price  level"1:  for  "trade  activity"  is  obviously  not 
unconnected  with  the  variations  in  deposits,  which,  according  to  Mr. 
Snyder,  are — along  with  the  "long-time  trend  of  trade-growth" — the 
only  factors  to  act  on  the  price  level.  The  picture  presented  by  English 
figures  (Chart  XXXV,  Chap.  XIII)  is  not  substantially  different. 
The  material  of  other  countries  does  not  lend  itself  to  a  similar  experi- 
ment, but  indications  may  be  had.  In  France,  for  instance,  the  "veloc- 
ity" of  the  net  (solde)  of  the  current  accounts  in  the  Banque  de  France, 
first  studied  by  P.  Des  Essars  (Journal  de  la  Societe  de  Statistique  de 
Paris,  April  1895),  gives  us  such  an  indication,  which  conforms  to 
expectation.2 

1  Cf.  Carl  Snyder,  The  Problem  of  Monetary  and  Economic  Stability,  Quarterly  Journal 
of  Economics,  February  1935,  p.  189.     Chart  IV,  p.  188,  may  be  referred  to  as  illustrating 
the  phenomenon  under  discussion.     But  it  must  be  born  in  mind  that  it  cannot  do  more 
than  give  a  very  rough  idea  of  it.     Outside  clearings  divided  by  Mr.  Snyder's  index  of  the 
general  level  of  prices,  corrected  for  trend  and  seasonal  fluctuations,  stand  up  to  1919  for 
output.     And  total  clearings  divided  by  national  bank  deposits  are  taken  to  represent 
"velocity"  or,  as  we  would  say,  rate  of  spending.     Neither  ratio  represents  satisfactorily 
what  it  should.     The  method  used  to  construct  the  output  series,  from  1919  on,  is  not  open 
to  the  same  objections.     But  the  behavior  of  the  other  series,  during  the  years  from  1926 
to  1930,  sufficiently  shows  that  it  cannot  reflect  adequately  the  rate  of  spending  in  the 
spheres  of  consumption  and  production.     Reference  should  be  made  to  Mr.  Snyder's  other 
papers  on  the  subject,  in  particular,  A  New  Index  of  Business  Activity,  Journal  of  the 
American  Statistical  Association,  March  1924;  Deposits  Activity  as  a  Measure  of  Business 
Activity,  Review  of  Economic  Statistics,  October  1924;  New  Measures  of  the  Relations  of 
Credit  and  Trade,  Academy  of  Political  Science,  Proceedings,  January  1930,  to  which  is 
appended  a  bibliographical  note.     In  spite  of  all  that  we  may  have  to  urge  against  both 
the  statistical  finding  and  Mr.  Snyder's  interpretation  of  it,  the  facts  remain  that  the 
covariation  in  the  sense  outlined  in  the  text  between  output  and  "velocity"  is  definitely 
recognizable,  and  that  our  analysis  goes  some  way  toward  supporting  the  results  of  his. 
The  affinity  becomes  still  clearer  from  his  argument  on  p.  27  of  the  last-named  paper,  which 
explains  the  long-run  rate  of  increase  in  output  in  terms  of  what  we  call  growth  and  evo- 
lution without,  however,  connecting  the  latter  with  cyclical  fluctuations.     It  is  interesting 
to  compare  Mr.  Snyder's  investigation  with  Mr.  Holbrook  Working's  study  on  the  relation 
between  fluctuations  in  deposits  and  in  the  price  level  (Bank  Deposits  as  a  Forecaster  of 
the  General  Level  of  Wholesale  Prices,  Review  of  Economic  Statistics,  1925).     Mr.  Working 
finds  fairly  close  covariation  (somewhat  improved  by  lagging)  between  the  latter  and  the 
deviations  of  deposits  from  a  trend  of  growth,  which  may  be  roughly  identified  with  the 
trend  in  output.     He  does  not  stress  the  cyclical  variability  of  "velocity."     But  since 
output  as  a  matter  of  fact  fluctuates  cyclically,  his  result,  such  as  it  is,  implies  that.     See, 
also,  Mr.  Snyder's  charts  on  pp.  28  and  30  in  New  Measures  of  the  Relations  of  Credit  and 
Trade. 

2  See  G.  Roulleau,  Vitesse  de  Circulation  des  Diverses  Formes  du  Stock  Mon6taire, 
Journal  de  la  Societe  de  Statistique  de  Paris,  April  1937,  chart  on  p.  9.     In  order  to  appraise 
the  significance  of  these  figures,  it  is  necessary  to  keep  in  mind  the  peculiar  position  of  the 
Banque  de  France  in  the  French  economy. 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES      587 

4.  Underspending  is  particularly  in  evidence  in  those  rare  cases 
in  which  deposits  decrease  absolutely:  system  expenditure  as  indicated 
by  outside  clearings  then  decreases  still  more  (observe  the  declines  to 
1885,  to  1894,  and  to  1908;  if  deposits,  see  American  pulse  chart  (VI), 
increased  in  1874  and  1875,  this  was  presumably  due  to  the  fact  that  the 
national  banking  system  gained  ground  and  to  the  immigration  of  currency 
into  banks).  Such  "hoarding"  is  wholly  consequential  and,  as  has  been 
pointed  out  above  (1,  a),  fundamentally  the  same  phenomenon  as 
shrinkage:  if  business  were  entirely  financed  by  bank  loans  and  if  there 
were  no  other  deposits  than  those  created  by  these  loans,  there  would 
be  no  underspending,1  but  all  its  consequences  would  still  be  present 
because  of  the  additional  shrinkage  which  then  would  take  the  place 
of  underspending.  It  does  not  follow  that  measures  directed  against  it 
are  meaningless.  Both  underspending  and  nonborrowing  are  though 
consequential  yet  sufficiently  important  in  themselves  and  as  "causes 
of  secondary  effects"  to  warrant  operating  on  them  directly.  Barring 
questions  of  technique,  the  problem  is  the  same  in  both  cases.  Buy-now 
campaigns,  public  works,  even  stamped  money,  and  so  on  may  help  to 
break  spirals  and  to  relieve  depressive  situations.  Such  treatment  is, 
to  be  sure,  merely  dermatological.  It  also  may  have,  and  in  general  has, 
other  effects  besides  the  one  desired.  But  it  is  not  futile. 

This,  however,  has  little  to  do  with  Saving.  Inasmuch  as  the  latter 
presupposes  the  existence  of  owned  balances  (including  cash)  which  do 
not,  except  in  the  case  of  bankruptcy,  disappear  when  not  in  use,  it  will 
indeed  tend  to  accentuate  underspending  as  against  shrinkage  of  deposits. 
Since  it  is  immaterial  which  happens,  there  is  no  point  in  trying  to  stimu- 
late expenditure  by  penalizing  Saving  except  this:  as  far  as  saving  is — 
though  not  necessarily — associated  with  real  investment  and  real  invest- 
ment with  postponable  expenditure  on  durable  producers'  goods,  savings 
are,  in  fact,  more  likely  to  become  idle  than  the  sums  earmarked  for 
expenditure  on  transient  consumers'  goods  of  the  "necessary"  type. 
This  consideration  does  not,  of  course,  suffice  to  "justify"  a  long-run 
policy  that  aims  at  transforming  the  former  into  the  latter,  which, 
if  embarked  upon  a  century  ago,  would  have  made  it  practically  impossi- 
ble to  attain  the  present  standard  of  life  of  the  masses  and  which  would 
logically  also  have  to  be  applied  to  sums  intended  for  expenditure  on 
durable  consumers'  goods.  Moreover,  recalling  our  analysis  in  Chap. 
Ill,  Sec.  A,  and  our  discussion  of  individual  crises  and  depressions  in 

1  To  be  exact,  we  ought  to  say  that  there  will  be  less  of  it;  for  some  decrease  in  activity 
will  occur  also  with  borrowed  deposits.  Moreover,  there  is  such  a  thing  as  borrowing  in 
order  to  be  liquid.  But  its  importance  can  hardly  be  great,  and  it  is  much  more  likely  to 
occur  in  prosperity — when  opportunities  may  be  expected  to  present  themselves  suddenly — 
than  in  depression,  when  paying  off  bank  debts  or  not  renewing  loans  is  the  most  orthodox 
thing  to  do. 


588  BUSINESS  CYCLES 

Chaps.  VI  and  VII,  we  conclude  that  depressions  may  be  expected  to 
be  the  milder,  the  more,  other  things  being  equal,  the  expenditure  of 
preceding  prosperities  has  been  financed  by  saving  and  accumulation. 
But  again,  as  a  temporary  expedient  to  be  applied  when  depression  has 
already  set  in,  such  an  antisaving  attitude,  though  still  more  productive 
of  undesired  consequences  than  simple  stimulation  of  spending,  cannot 
be  called  futile.1 

Of  course,  it  would  change  but  the  form  and  not  the  §ubstance  of  our 
argument  if  we  defined  saving  and  accumulation  so  as  to  include,  or  to 
coincide  with,  nonspending.  Tautological  propositions  that  would  then 
follow  about  the  effects  of  saving  in  this  sense  cannot  teach  anything 
about  the  effects  of  thrift.  That  argument  does  not,  however,  dispose 
of  the  question  what  these  effects  actually  are.  One  point  has  to  be 
added,  which  turns  on  the  relation  between  saving  and  credit  creation: 
although  in  a  system  that  is  beyond  the  stage  of  what  we  have  called 
the  immigration  of  money  into  banks,  "savings  do  not  create  deposits," 
they  annihilate  deposits  whenever  they  are  applied  to  the  repayment  of 
bank  loans.  This  is  in  fact  one  of  the  major  pieces  of  our  mechanism — 
repayments  out  of  profits  induce  autodeflation,  or,  to  put  it  differently, 
accumulating  profits  ex  post  finance  the  real  investment,  which  it  was 
necessary  to  effect  in  order  to  reap  them.  If,  in  addition  to  this,  the 
savings  of  households  are  borrowed  for  the  same  purpose,  the  process  is, 
of  course,  accelerated  thereby.  Savings  thus  step  in  to  relieve  bank 
credit,  and  in  this  sense  the  old  theory  that  it  is  savings  that  finance 
the  expansion  of  the  industrial  apparatus  comes  partly  true  after  all, 
even  in  the  presence  of  credit  creation.  But  they  do  so  with  a  lag, 
which  is  responsible  for  a  sequence  of  phenomena  that  would  be  absent2 
if,  instead  of  stepping  in  at  a  later  stage,  savings  financed  enterprise 
from  the  start,  and  that  would  be  less  in  evidence  if  they  did  not  step 
in  at  all.  We  are  obviously  dealing  with  a  very  typical  practice. 
Nothing  is  more  usual,  both  in  the  case  of  enterprise  in  our  sense  and 
in  the  case  of  induced  expansion,  than  to  go  ahead  on  bank  credit  and 
then  to  "fund"  the  debt  by  the  issue  of  stocks  and  bonds.  It  is  in  this 
role,  rather  than  in  the  role  of  primary  source  of  means,  that  we  would 
have  introduced  saving  into  our  pure  model,  had  we  introduced  it  at 
all. 

This  deflationary  effect  of — not  saving  and  accumulation  as  such 
but — the  application  of  savings  and  accumulations  to  the  repayment  of 
bank  loans  is  what  can  be  adduced  in  support  of  oversaving  theories  and 

1  What  contradiction  may  possibly  be  suspected  to  exist  between  the  last  two  sentences 
of  this  paragraph  will  presently  disappear. 

2  By  this,  it  will  be  recalled,  we  do  not  mean  the  cycle  itself  but  certain  features  which, 
though  secondary  in  logic,  yet  are  of  prime  importance  practically,  for  instance,  fluctuations 
of  the  price  level  which  superimpose  themselves  on  and  accentuate  the  fundamental  ones. 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES      589 

antisaving  policies.  Neither  the  effects  of  the  saving  and  investment 
process  within  the  monetary  or  the  commodity  sphere,  nor  such  cyclical 
variability  as  may  with  justice  be  attributed  to  it — the  "cyclical  release 
of  saving" — can  offer  any  substantial  standing  ground  for  them.  But 
it  may  indeed  be  urged  that,  as  far  as  savings  destroy  deposits  which  would 
not  become  idle,  they  in  fact  intensify  recession  and  depression.  The 
losses  incident  to  these  processes  will  then  be  greater  than  they  would 
be  if  the  same  sums  had  been  spent  on  consumers'  goods.  And  it  might 
even  be  held  that,  thus  applied,  savings  do  not  serve  any  social  purpose 
because,  once  the  task  which  older  theory  assigned  to  them  is  fulfilled 
by  credit  creation,  they  seem  to  come  in,  as  it  were,  for  no  other  purpose 
than  to  create  trouble  additional  to  that  which  innovation  would  create 
in  any  case.  Here  we  come  nearer,  therefore,  to  the  standpoint  of 
oversaving  theories — although  it  is  not  any  "excess"  of  saving  that  is 
responsible  for  this  additional  trouble — than  the  general  drift  of  our 
argument  may  have  led  the  reader  to  expect. 

The  case  which  can  be  made  out  on  these  lines  is,  however,  much 
weaker1  than  it  seems.  For,  although  there  is  no  way  of  determining 
exactly  the  quantitative  importance  of  this  particular  use  of  household 
savings,  the  behavior  of  our  series  shows  that  it  can  hardly  ever  have  been 
responsible  for  absolute  fall  in  deposits  or  in  the  sum  total  of  incomes; 
and  only  if  it  had  been,  would  it  have  had  really  serious  effects.  This  is 
due  to  the  fact  that,  as  a  rule,  especially  during  prosperity  but  also  in 
recession,  repayment  from  savers'  funds  merely  serves  to  set  free  banks' 
facilities  for  lending,  which  are  then  promptly  used  by  other  borrowers; 
and  therefore  it  seems  to  contribute  but  little  to  the  coming  about  of 
the  upper  turning  point,  except  in  price  level  and  interest  rates.  More- 
over, a  stabilizing  effect  of  the  practice  must  not  be  forgotten.  On 
the  one  hand,  firms  which  succeed  in  substituting  bonds  or  shares  to 
their  debt  at  their  banks  thereby  consolidate  their  position  and  thus 
restrict  the  extent  of  the  danger  zone.  On  the  other  hand,  as  far  as  this 
is  done  during  prosperity,  it  tends  to  bridle  excesses  and  thus  to  mitigate, 
rather  than  accentuate,  reactions.  It  is  different,  of  course,  in  recession 
and  as  regards  the  danger  that  recession  slide  off  into  a  spiral,  when  well- 
timed  and  well-dosed  government  expenditure  could  no  doubt  do  much 
to  counteract  what  then  easily  turns  into  functionless  catastrophe. 

A  complete  theory  of  saving  and  accumulation  could,  so  the  writer 
believes,  be  constructed  from  the  disjecta  membra  that  the  reader  finds 
in  almost  every  chapter  of  this  book.  But  he  frankly  admits  that  its 

1  Since  analysis  is  our  only  aim,  there  is  no  need  to  go  into  the  question  as  to  how  far 
antisaving  recommendations,  given  certain  aims,  say,  of  stabilization,  may  be  said  to  follow 
from  the  above.  Taking  the  practical  question  in  all  its  bearing,  the  writer  personally 
thinks  that  there  is  no  case  at  all  for  them.  But  that  is  for  the  reader  to  decide  according 
to  his  valuations  and  according  to  his  confidence  in  agencies  of  control. 


590  BUSINESS  CYCLES 

factual  complement  is  far  from  satisfactory.  This  is  primarily  due  to  a 
lack  of  data,  which,  for  prewar  times  at  least,  confronts  any  student  of 
the  subject,  however  he  may  arrange  his  terminology.  As  far  as  it  is 
also  due  to  the  particular  definition  of  saving  and  accumulation  adopted 
for  the  purposes  of  this  book,  the  obvious  reply  to  possible  objections 
is  that  there  was  no  choice.  Nothing  would  have  been  gained  by  the 
use  of  a  concept  that,  while  lending  itself  a  little  better  to  statistical 
evaluation,  would  be  irrelevant  to  the  analysis  of  the  real  processes 
at  work :  the  insoluble  problem  of  how  to  extricate  the  relevant  elements 
would  arise  nevertheless.  We  are,  for  instance,  somewhat  better  off 
with  respect  to  real  investment  ("capital  formation")  and  to  investment 
in  general,  although  even  these  figures  cannot  be  trusted  beyond  what  we 
know  independently  of  them.1  But  we  cannot  improve  our  situation  by 

1  This  must  be  obvious  to  anyone  who  undertakes  a  critical  study  of  what  is  still  the 
standard  performance  for  England  and  the  greater  part  of  the  nineteenth  century  (especially 
1865—1885;  there  is,  however,  a  retrospect  that  notices  the  more  important  attempts  back 
to  Petty,  Davenant,  and  King):  R.  Giffen's  Growth  of  Capital,  1889.  The  estimates, 
particularly  of  what  he  calls  accumulation  and,  within  accumulation,  "free  saving"  (i.e., 
the'  saving  that  is  invested  through  the  stock  exchange),  are,  even  though  reasonable, 
much  too  rough  to  be  of  use.  Later  on,  we  have  interesting  figures  for  particular  years, 
e.g.,  from  the  Census  of  Production,  1907,  and  also  series  for  individual  items.  They  did 
not  seem  to  warrant  an  attempt  at  constructing  a  general  series,  which  the  writer  hopes, 
however,  will  be  materially  facilitated  by  a  forthcoming  work  by  Mr.  Cairncross. 
No  estimate  of  yearly  "capital  formation"  seems  possible  for  Germany,  but  interesting 
samples  may  be  culled  from  the  behavior  of  companies,  one  of  which  will  be  mentioned 
later.  To  a  certain  extent,  this  can  be  done  also  for  this  country,  but  there  is  besides  a 
source  that  is  peculiar  to  it  and  has  recently  been  exploited  by  Mr.  Snyder.  (See  Capital 
Supply  and  National  Well-being,  American  Economic  Review,  June  1936;  also  Die  Bedeu- 
tung  des  Kapitalangebots  ftir  den  Industrialisierungsprozess,  Weltwirtschaftlich.es  Archiv, 
September  1935.)  It  is  the  Census  Bureau's  data  beginning  with  1826  on  capital  invested 
in  manufacturing — value  of  plant,  cash,  inventories,  receivables,  and  so  on.  The  writer 
has  not  been  able  to  satisfy  himself  whether  value  of  plant  means  original  cost,  replacement 
cost,  or  book  value,  and  is  also  on  other  grounds  disposed  to  believe  that  the  bureau  dis- 
continued publication  of  this  information  not  entirely  without  reason  (see  the  remarks  in 
the  Xllth  Census  of  Manufactures,  Part  I,  p.  xcvi).  The  relation  of  the  "value  of  real 
capital"  to  domestic  savings  must,  moreover,  be  expected  to  be  a  particularly  distant  one  in 
the  case  of  the  United  States.  Yet  he  gratefully  recognizes  the  merit  of  Mr.  Snyder's 
bold  step.  The  relation  of  the  curve  to  that  of  National  Income  (see  op.  cit.t  chart  on 
p.  203)  is  certainly  what  a  theorist  would  expect.  The  figures  for  1889,  1899,  and  1904, 
excluding  working  capital,  have  been  ingeniously  used  by  Professors  Douglas  and  Cobb 
(American  Economic  Review,  March  1928,  Supplement,  p.  139)  as  stepping  stones  for  an 
interpolation — although  they  are  not  comparable,  both  because  they  do  not  each  time 
include  the  same  things  and  because  the  mergers  greatly  influenced  valuations — by  means 
of  an  index  of  the  quantities  and  prices  of  the  commodities  chiefly  entering  into  capital 
goods  so  as  to  yield  yearly  figures.  Critical  comments  crowd  upon  us  (for  some  of  them, 
see  Professor  Slichter's  remarks  in  the  discussion  of  the  paper),  but  results  may  yet  be 
nearer  to  the  true  contours  than  one  may  feel  justified  in  hoping.  We  shall  meet  the 
subject  again,  but  until  we  reach  the  postwar  period  shall  continue  to  use  only  those  indi- 
cators which  have  already  been  used  in  the  preceding  sections. 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES      591 

defining  saving  so  as  to  make  it  equal  to  or  identical  with  investing;  for 
this  would  simply  mean  begging  the  most  interesting  question  that  arises 
in  this  connection,  because  it  is  precisely  the  relation  between  those  two 
distinct  processes  on  which  more  light  would  be  desirable.  A  similar 
objection  estops  us  from  accepting  the  amount  of  new  security  issues 
taken  by  the  public  as  an  indicator  of  saving,  which,  moreover,  is  vitiated 
by  foreign  buying  and  the  fact  that  to  a  great  and  varying  extent  sub- 
scriptions are  financed  by  bank  loans. 

Figures  for  corporate  accumulation  are  for  prewar  times  available 
only  in  samples,  the  value  of  which  is  reduced  by  the  dependence  of  sur- 
plus on  a  largely  arbitrary  variable — depreciation — and  by  revaluation 
of  assets.1  Estimates  of  accumulations  by  farmers,  craftsmen,  and 
private  firms  in  general,  which  are  ploughed  back  into  their  own  business, 
must  be  highly  unreliable.  The  greatest  disappointment,  however, 
awaits  those  who  still  cling  to  the  idea  that  at  least  deposits  not  subject 
to  check  are,  ipso  facto,  savings.  This  is  most  nearly  true  of  the  deposits 
in  mutual  and  stock  savings  banks  in  this  country  and  of  the  Sparkassen 
in  Germany.  But  even  neglecting  the  fact  that  the  Continental  saving 
institution  also  serves  as  the  bank  of  the  small  man  and  that,  as  far  as  it 
does,  increase  in  its  deposits  indicates  not  saving  but  simply  transition 
to  another  method  of  keeping  cash,  we  must  not  forget  that  a  great 
part  of  those  funds  is  assembled  for  definite  acts  of  expenditure — a 

1  More  ambitious  attempts  would  be  possible,  at  least  for  Germany,  if  one  could 
neglect  those  difficulties  and  a  few  others.  A  very  interesting  study  made  by  the  Institut 
flir  Konjunkturforschung  (Anlagetaetigkeit  und  langfristige  Finanzierung  in  der  Industrie 
vor  dem  Kriege,  Vierteljahrshefte  zur  Konjunkturforschung,  1929),  which  undertakes  to 
overcome  them,  must,  however,  be  mentioned.  It  deals  with  20  industrial  companies 
only  (capital  in  1918,  404  million  marks),  some  of  them  quite  young,  for  1896  to  1913. 
Accumulation  is  put  equal  to  the  sums  "reserved"  (out  of  net  profits),  plus  the  sums 
actually  written  off,  minus  "technologically  necessary"  depreciation.  This  "necessary" 
depreciation  is,  of  course,  not  above  objection,  nor  is  the  method  by  which  it  was  arrived  at, 
if  we  may  judge  from  the  short  sentence  describing  it;  but  results  are  hardly  invalidated 
thereby.  We  learn  that  accumulation  (innere  Kapitalbildung)  varied  much  as  dividends 
did,  though  much  more  strongly;  that  in  the  average  of  the  sample  it  varied  cyclically 
between  0.6  and  4.8  per  cent  of  the  paid-up  capital;  that  in  the  average  of  years  (between 
1900  and  1913)  it  was  2.5  per  cent  of  the  paid-up  capital,  which  equates  it  to  about  20  per 
cent  of  the  total  net  surplus  over  cost,  excluding  necessary  depreciation,  tantiemes,  social 
outlay  (expenditure  from  net  profits  for  the  benefit  of  workmen),  and  sums  carried  forward 
(i.e.,  of  the  total  of  disbursements  to  shareholders  plus  accumulation);  that  these  com- 
panies on  the  average  raised  fresh  long-term  funds — by  the  issue  of  shares  and  bonds — to 
the  yearly  amount  of  7.7  per  cent  of  their  paid-up  capital  (as  it  stood  at  the  beginning  of 
each  year);  and  that  more  than  all  of  these  means  was  regularly  invested  in  plant  and 
equipment  before  the  end  of  each  prosperity :  investing  outran  long-time  financing,  occasion- 
ally even  in  depressed  years.  It  must  be  remembered  that  the  concerns  investigated  were 
highly  progressive  and,  as  said  before,  in  part  relatively  new  ones.  The  sample  is,  hence, 
not  representative.  But  it  suffices  to  correct  many  a  mistaken  idea  about  the  invest- 
ment process. 


BUSINESS  CYCLES 

holiday,  marriage,  a  cottage,  and  so  on,  or  less  definitely  for,  say,  illness 
or  a  rainy  day.  Again,  the  reader  may  think  as  he  pleases  about  the 
appropriateness  of  the  writer's  terminology,  which  excludes  from  savings 
these  sums  intended  to  be  spent.  But  if  he  includes  them,  he  will  have 
nevertheless  to  recognize  the  fact  that  the  sum  total  of  these  "savings" 
is  currently  drawn  upon  for  purposes  of  consumption  by  the  "savers" 
themselves,  which  makes  all  the  difference.1  This  applies  still  more  to 
the  time  deposits  in  commercial  banks,  only  a  fraction  of  which  can 
possibly  be  looked  upon  as  genuine  savings.  The  rest  represents  under- 
spending, temporary  investment,  and,  again,  preparation  for  "bulky" 
acts  of  expenditure,  while,  as  has  been  pointed  out  before,  a  part  of  the 
total  does  not  differ  in  nature  from  demand  deposits.  Finally,  the 
method  of  deriving  households'  savings  by  deducting  expenditure  on 
consumers*  goods  from  income2  minus  taxes  and  charities,  is  not  available 
for  the  prewar,  scarcely  even  for  the  postwar,  epoch.  But  even  if  it 
were,  we  should  have  to  remember  that  the  presence  of  acts  of  expenditure 
which  recur  in  periods  longer  than  the  income  period  must  seriously 
impair  the  significance  of  the  results.  In  an  investigation  into  the 
effects  of  saving  on  the  economic  process  and,  incidentally,  into  the 
validity  of  oversaving  theories  of  the  business  cycle,  it  would  obviously 
be  absurd  to  include  in  savings  elements  of  income  earmarked  for  expendi- 
ture on  durable  goods,  such  as  houses,  furniture,  and  motorcars.  Yet 
this  absurdity  is  not  only  difficult  to  avoid,  but  its  influence  must  be 
expected  to  increase  with  time,  since  in  a  progressive  community  those 
elements  of  income  will,  in  general,  gain  in  relative  as  well  as  absolute 
importance. 

Certain  facts  are  nevertheless  quite  clear.  The  most  important  of 
them  is  that  the  amount  of  saving  in  our  sense — and,  be  it  repeated, 
producing  a  larger  amount  by  means  of  a  wider  definition  would  not  alter 
the  situations,  since  the  additional  elements  then  included  would  not 
have  the  same  effects — must  be  much  smaller  than  is  commonly  supposed. 
For  with  a  possible  exception  on  the  score  of  hidden  reserves  of  firms, 
all  our  criticisms  of  the  data  tend  to  reduce  it.  It  is  also  clear — in  fact, 
a  matter  of  common  experience — that  the  behavior  of  firms'  accumulation 

1  Similar  considerations  also  apply  to  part  of  the  assets  of  building  and  loan  associations 
and  of  life  insurance  companies,  which  reflect  methods  of  providing  for  future  expenditure 
that  are  alternative  to  saving. 

2  It  will  be  remembered  that  income  in  our  sense  excludes  capital  gains.     Since  they 
emerge  outside  of  the  flow  of  system  expenditure,  there  is  for  purposes  of  business-cycle 
study  no  justification  for  including  them.     This  has,  however,  been  done  in  the  recent 
investigation  of  the  Brookings  Institution,  and  it  is  to  this  item  that  both  the  fantastically 
high  estimate  arrived  at  and  the  rate  of  increase  of  "saving"  in  the  later  twenties  of  this 
century  are  due.     Spending  on  consumers'  goods  of  such  gains  is,  of  course,  dis-saving  and 
is  on  a  par  with  financing  consumers'  expenditure  by  borrowing.     See  Chap.  XIV. 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES       593 

must  be  strongly  cyclical.  Dividends  do  not  in  general  fluctuate  as 
much  as  does  net  revenue.  But  since  we  exclude  capital  gains,  the 
behavior  of  households'  saving  is  more  doubtful,  because  the  opportunity 
to  save  afforded  by  an  increase  in  real  income  may  be  counteracted  by 
optimistic  anticipations,  as  may  the  effect  of  a  decrease  by  pessimistic 
anticipations.  It  also  varies  greatly  with  groups,  countries,  times 
(see,  above,  Sec.  B).  But  such  indications  as  we  have  strongly  point 
toward  steadiness  of  rate.  This  is  the  result  which  has  been  arrived  at 
for  postwar  Germany  by  Professor  Wagemann,1  and  which  suggests 
itself  for  this  country  also,  if  we  may  put  our  trust  in  the  evolution  from 
1880  of  the  assets  of  life  insurance  companies2  and  in  the  deposits  in 
mutual  and  stock  savings  banks.  The  latter  display  a  (logarithmically) 
straight -line  trend  from  1840  to  1876 — which,  no  doubt,  is  largely  a 
"special"  one  in  our  sense  of  this  term — the  small  deviations  from  which 
are,  however,  clearly  positive  during  the  prosperities  of  the  fifties  and 
early  seventies.  Then  we  observe  decline  to  1879  and,  from  that  year 
on,  again  straight-line  increase  (though  of  a  smaller  gradient),  inter- 
rupted by  kinks  in  1893,  1908,  and  1914.  So  depression,  as  well  as 
prosperity,  does  assert  itself  in  this  series,  but,  the  material  being  what 
it  is,  the  outstanding  fact — the  steadiness — is  the  only  safe  one  to 
stress. 

Again,  as  in  the  case  of  the  application  of  savings  to  the  repayment  of 
bank  loans,  it  should  be  noticed  that,  even  if  saving  and  accumulation 
involved  the  "locking  up  of  money/'  the  consequences  usually  attributed 
to  them  by  antisaving  theories  would  not  necessarily  follow.  In  that 
case,  accumulation  from  profit  would  tend  to  dampen  the  excesses  of 
booms  exactly  as  restrictive  open-market  operations  undertaken  by 
central  banks.  And  the  dis-saving  during  deep  depressions — the 
depletion  of  the  accumulated  surpluses  of  corporations  by  deficits  and 
dividend  payments  and  of  the  saving  deposits  of  the  small  savers  by 
current  expenditure — would  tend  to  alleviate  them.  If  we  observe  so 
little  of  these  equilibrating  effects,  this  is  because  there  is  so  little  "locking 
up."  Also,  the  importance  of  the  cyclical  "release"  of  savings  should 
not  be  overrated.  Prosperities  are,  to  be  sure,  periods  of  supernormal 
real  investment;  and  depressions,  of  subnormal  real  investment.  Savings 
like  other  funds  are  active  in  the  former  and  inactive  in  the  latter.  But, 
as  we  have  seen,  it  does  not  follow  that  investment  must  decrease  in 
recession,  for  all  the  demand  for  funds  incident  to  the  conquest  of  new 
economic  space  then  asserts  itself,  and  there  is,  moreover,  that  type  of 

1  See,  for  example,  Einfuehrung  in  die  Konjunkturlehre,  1929,  p.  117. 

2  See  Carl  Snyder,  op.  cit.,  chart  on  p.  203.     That  evolution  is  almost  perfectly  fitted 
by  a  (logarithmic)  straight  line  and  displays  but  insignificant  fluctuations.     Observe  the 
relation  almost  amounting  to  parallelism  with  Capital  Invested  in  Manufactures. 


594 


BUSINESS  CYCLES 


demand  which  waits  upon  a  fall  in  interest — demand  for  dwelling-house 
building,  in  particular. 

5.  We  return  again  to  Chart  XXV,  in  order  to  study  more  closely  the 
relation  between  customers'  balances  and  loans  (and  discounts)  of 
national  banks.  As  has  been  pointed  out  in  2,  the  picture  "is  of  two 
series  which  move  closely  together,  and  yet  with  notable  differences" 
(A.  A.  Young,  op.  cit.9  p.  5).  The  covariation  stands  out  particularly 
well  if  we  eliminate  the  descriptive  trend,  an  operation  that  is  less 
objectionable  than  it  otherwise  would  be,  if  we  keep  within  a  Kondratieff 


.OANS  AND  DISCOUNTS 


TOTAL  DEPOSITS 


NOTE  CIRCULATION 


1901        1902       1903        1904       1905        1906       1907        1908       1909       1910        1911 
CHART  XXXII.  —  Germany  (see  Appendix,  p.  1063). 


1912        1913 


phase,  as  Professor  Young  substantially  does  in  his  chart  15  F  (ibid., 
p.  27),  l  which  covers  the  years  from  1901  to  1914.  But  the  essential 
fact  so  important  for  the  understanding  of  prewar  banking  at  least  — 
viz.j  that  the  movement  of  loans  dominates  the  movement  of  deposits  — 
is  obvious  in  any  case.  Nor  is  this  covariation  confined  to  those  shorter 
fluctuations  which  are  traditionally  recognized  as  "cycles."  Of  course 
the  longer  the  period  we  survey,  the  more  extraneous  elements  will 
assert  themselves,  and  we  shall  not  be  surprised  to  find  that  covariation 
is  particularly  close  in  the  course  of  Kitchins  and  Juglars.  But  the 
phases  of  the  two  Kondratieffs  are  also  recognizable  and  influence  both 
series  similarly.  Chart  XXXII  presents  the  German  case.2 

1  Comparison  is  there  with  Net  Deposits,  but  this  does  not  matter  for  cyclical,  although 
it  would  for  seasonal,  movements;  cf.,  same  chart  B. 

*  In  perusing  that  chart,  it  is  necessary  to  keep  in  mind  that  total  deposits  do  not  mean 
exactly  the  same  thing  or  hold,  as  far  as  they  do  mean  a  comparable  magnitude,  the  same 
relative  position  as  in  the  United  States.  It  must  also  be  remembered  that  deposit  banking 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES      595 

The  covariation  between  loans  and  deposits  must  of  course  be  dis- 
turbed by  member  banks*  investments.  In  national  banks  outside 
New  York  City  they  fell  from  a  level  of  roughly  10  per  cent  of  loans  and 
discounts  at  the  beginning  of  the  series  to  a  minimum  of  24  millions 
(Feb.  28,  1873),  but  within  our  period  they  reached  898  millions,  or 
about  one-sixth  of  loans  and  discounts  (June  30,  1914),  the  increase  from 
1897  to  1902  helping  to  push  the  deposit  up  to  the  loan  series.  This 
subject  belongs  to  the  sphere  of  banking  practice  and  will  be  taken  up 
in  Chap.  XIII;  but  we  may  note  at  once  that  the  impression  of  an  erratic 
behavior  of  investments  in  the  course  of  cycles  is,  apart  from  the  special 
trend  that  is  clearly  present  in  them,1  due  to  the  conflict  of  two  tendencies : 
the  one  which  clearly  prevails,  at  least  in  the  shorter  fluctuations  (see 
A.  A.  Young's  chart  15  H,  op.  cit.,  p.  27),  and  works  toward  negative 
association  of  investments  with  loans,  the  banks  employing  in  slack 
times  idle  funds  in  the  purchase  of  assets,  thereby  creating  idle  deposits 
that  help  to  steady  the  deposit  series ;  and  another  and  weaker  one,  which 
proceeds  from  firms  and  individuals'  liquidating  their  holdings  of  bonds — 
when  in  busy  times  they  have  more  profitable  uses  for  their  funds — 
and  which  works  towards  positive  association  with  loans  and  intensifies 
fluctuations  of  deposits. 

But  there  are  other  factors  which  will  enforce  deviation  from  the 
fundamental  parallelism  of  loans  and  balances.  One  is  variation  in  the 
capital  and  surplus  item.  In  national  banks  outside  New  York  City 
this  item  grew  from  less  than  400  millions  in  1867  to  over  1^  billions 
in  1914.  It  clearly  displays  the  influence  of  the  cyclical  process,  following 
as  it  does  on  a  trend  of  smaller  gradient2  the  movement  of  earning  assets 
(see  A.  A.  Young,  op.  cit.,  Chart  2  on  p.  5  and  Chart  9  on  p.  22).  Another 
factor  is  variation  in  original  deposits.  The  spread  of  banking  habits 
and  the  consequent  immigration  into  banks  of  money  that  used  to 
circulate  outside  of  them  would  in  itself  suffice  to  impart  another  special 
trend  to  the  loan-deposit  ratio.  Apart  from  this,  it  is  easy  to  realize 
that  original  deposits  will  also  fluctuate  cyclically  and  that  they  will 
reflect  the  influence  of  gold  movements,  although  in  the  latter  respect 
it  is  interesting  to  note  that  neither  the  deposit  nor  the  loan  series  behaves 
very  differently  before  and  after  the  intrusion  of  the  new  gold.  If  we 
had  no  other  information  than  that  conveyed  by  those  two  curves, 

in  general  and  the  business  of  the  banks  that  contributed  the  figures  in  particular,  rapidly 
gained  ground  during  that  period.  This  partly  accounts  for  the  steep  gradient  from  1903 
to  1906  and  again  in  1910.  In  order  to  bring  this  out,  the  series  showing  the  development 
of  note  circulation  has  been  added. 

1  See  infra,  Chart  XXXIV. 

2  The  fact  that  that  gradient  is  smaller  than  the  gradient  of  the  earning  asset  curve 
reflects,  in  part,  the  relative  decrease  of  hand-to-hand  circulation  (see  below). 


596  BUSINESS  CYCLES 

we  should  hardly  be  able  to  infer  that  something  had  happened  to  alter 
the  monetary  data  of  our  process.  Finally,  as  long  as  coins  and  notes 
circulate  at  all,  and  especially  if  farmers,  craftsmen,  wage  earners,  and 
their  retailers  live  to  a  significant  extent  outside  the  banking  sphere,  the 
total  of  balances  will,  other  things  being  equal,  tend  to  increase  and 
decrease  cyclically  by  less  than  loans,1  because  cash  is  gradually  with- 
drawn from  every  original  or  newly  created  deposit  when  business  is 
brisk  and,  however  quickly  redeposited,  dwells  in  greater  amounts 
outside  the  banks  in  prosperity  than  in  depression. 

These  considerations  suffice  to  explain  why,  on  the  one  hand,  the  loan- 
deposit  ratio  is  itself  a  cyclical  variable,  and  why,  on  the  other  hand,  it 
is  not  always  easy  to  interpret  its  fluctuations  or  its  trend.  Mainly 
because  of  the  predominantly  inverse  association  between  investment  and 
loans  and  of  the  cyclical  drain  and  reflux  of  cash,  that  ratio  moves  with 
an  understandable  lag  in  the  same  direction  as  its  constituents.2  But 
several  circumstances  besides  those  mentioned  above — depletion  of 
deposits  by  hoarding  from  distrust  in  banks,  for  instance,  or  legislative 
changes — combine  to  interfere  with  that  rule.  We  may  now  inspect 
Chart  XXXIII,  on  which  circulation  (notes  outstanding)  has  been  added 
to  individual  deposits,  both  on  the  theory  that  bank  notes  are  funda- 
mentally the  same  kind  of  thing  as  customers'  balances  and  on  the  much 
more  doubtful  theory  that  the  variation  in  their  amount  is  indicative 
also  of  short  variations  in  the  amount  of  legal  tender  in  the  hands  of 
the  public,  therefore,  of  the  total  of  money  outside  the  Treasury  and 
all  banks.  Over  time,  of  course,  the  circulation  of  national  banks 
is  too  much  under  the  influence  of  factors  peculiar  to  them  to  be  of 
much  significance  for  us. 

As  far  as  the  variability  of  the  loan-deposit  ratio  is  due  to  the  cyclical 
depletion  and  repletion  of  deposits  by  cash  moving  out  and  in,  the  ratio 
between  money  in  circulation  and  deposits  must  also  be  a  cyclical  (and 
seasonal)  variable.  So  it  is,  of  course,  and  the  explicit  or  implicit 
assumption  characteristic  of  several  well-known  versions  of  the  quantity 
theory,  viz.,  that  it  is  constant,  can  only  hold  for  a  perfectly  stationary 
state.  In  a  nonstationary  system  this  assumption  cannot  even  hold 

1  Actually,  deposits  fluctuated  more  than  loans,  though  not  so  much  more  than  in  New 
York  banks.  This  is  not  difficult  to  understand.  It  must  be  remembered,  however,  that 
what  we  have  before  us  are  the  figures  not  of  a  complete  system  but  only  of  a  segment. 

8  According  to  Professor  W.  M.  Persons,  Cyclical  Fluctuations  of  the  Ratio  of  Bank 
Loans  to  Deposits,  1867-1914,  Review  of  Economic  Statistics,  October  1924,  this  lag  is 
about  six  months.  While  deposits  and  loans  themselves  fluctuate  concurrently  with  each 
other  and  the  Harvard  Business  Curve,  their  ratio  thus  fluctuates  concurrently  with  the 
Harvard  Money  Curve.  This  is  as  we  should  expect.  Professor  Persons,  however,  uses 
the  ratio  we  are  discussing  only  for  the  last  prewar  decade.  We  shall  not  be  astonished  to 
learn  that  those  relations  are  less  clearly  discernible  before  1900. 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES      597 

over  time — say,  for  successive  neighborhoods  of  equilibrium — for  our 
process  changes  the  relative  importance  of  the  transactions  directly 
settled  by  cash :  it  is  the  most  powerful  factor  responsible  for  the  immigra- 
tion of  money  into  banks,  i.e.,  the  spread  of  the  habit  of  paying  by  check, 
which  in  turn  is  the  most  powerful  factor  in  the  expansion  of  the  lending 


1870  1880  1890  1900  19)0    1913 

CHAHT  XXXIII.— United  States  (see  Appendix,  p.  1063).    - 

facilities  of  banks.  This,  first,  helps  to  account  for  the  fact  that  the 
loan-deposit  ratio  has  been  constantly  falling  over  time.1  While  this 
ratio  can  be  observed  in  the  (outside)  national  bank  data,  measurement 
of  the  second  ratio — that  between  currency  in  circulation  outside  the 

1  It  averaged  about  160  per  cent  in  the  seventies  and  fell  in  every  successive  decade, 
until  it  practically  reached  100  per  cent  by  1910.  This  is,  of  course,  not  only  due  to 
deposits  losing  less  and  less  to  outside  circulation,  nor  can  the  sharp  jerk  from  1879  to  1881 
be  explained  in  this  way.  A  few  figures  should  be  added:  Between  1875  and  1914  loans  and 
discount  rose  about  seven  fold,  individual  deposits  more  than  elevenfold  (11.8).  "Lawful 
Money  Held"  by  the  same  banks  was  76  millions  on  Mar.  1,  1875,  and  642  millions  on 
Mar.  4,  1914;  Capital  and  Surplus  on  the  same  dates  were  respectively  536  and  1,539. 


598  BUSINESS  CYCLES 

Treasury  and  all  banks  and  total  deposits — is,  as  we  have  seen,  a  more 
doubtful  matter.  But  at  least  from  1890  on,  these  data  cannot  be  so 
faulty  as  to  shake  our  confidence  in  the  strongly  falling  trend  they 
display.1  Third,  the  ratio  of  outside  circulation  to  loans  plus  investments 
also  has  an  obvious  significance  of  its  own.  Mr.  Carl  Snyder  tried  to 
estimate  it2  from  an  estimated  series  of  outside  money  and  from  national 
bank  loans  and  discounts  "adjusted  upward  to  the  level  of  all  commercial 
banks  in  1913."  But  whatever  might  be  urged  against  this,  the  salient 
fact  is  again  beyond  doubt.  Since  the  fall  in  all  three  ratios  is  sub- 
stantially the  result  of  the  cyclical  process  of  evolution,  we  may  speak  of 
result  trends. 

6.  The  reader  should  once  more  refer  to  Chart  XXV  and  then  inspect 
Chart  XXXIV,  which,  although  very  inadequately,  presents  the  behavior 
of  indicators  of  what  is,  in  one  sense  or  another,  called  Investment.  The 
graph  of  member  banks'  investments  as  represented  by  the  investments 
of  national  banks  outside  New  York  City  has  been  inserted  merely  for  the 
purpose  of  showing — what  would  be  more  obvious  if  a  descriptive  trend 
had  been  eliminated — that  their  fluctuations  are  about  as  much  inversely 
related  to  production  of  industrial  equipment  as  they  are  to  loans  and 
discounts,  that  is  to  say,  predominantly  yet  not  entirely  so.  There  is 
again  more  than  a  suggestion,  as  there  was  in  the  comparison  of  loans 
with  deposits,  to  the  effect  that  selling  assets  to  banks  is  to  a  significant 
extent  resorted  to  by  individuals  and  firms  as  a  method  of  providing 
funds  for  industrial  expansion.  Building  Permits  do  not  merely  reflect 
investment — since  the  building  of  residences  for  one's  own  use  is  not 
investment  in  our  sense — and  behave  as  erratically  as  such  a  composite 
of  consumers'  and  investment  goods  should.3  All  affinity  with  the  move- 

1  C/.  Professor  Angell's  ratio  h  on  his  Chart  II  in  Behavior  of  Money,  1936.     The  deposit 
figures  used  include  all  deposits  in  reporting  banks,  also  in  postal  savings  banks,  and  both 
government  and  interbank  deposits.     The  float  is,  however,  deducted.     But  the  present 
writer  is  not  able  to  agree  with  his  statements  that  outside  currency,  circulating  deposits 
(deposits  subject  to  check — bankers'  balances — clearing  house  exchanges)  and  the  ratios 
of  outside  currency  and  circulating,  as  well  as  total,  deposits  do  not  display  clear  cycles  or 
that  there  is  no  evidence  of  direct  causal  connection  between  currency  and  deposits.     On 
the  contrary,  all  our  fluctuations  seem  to  show  very  well:  the  downward  sweep  of  ratio  h 
from  24  per  cent  in  1890  to  10  per  cent  in  1914  is  broken  in  every  prosperity,  intensified  in 
every  recession  or  depression.     It  should  be  added  that  that  downward  tendency  is 
exaggerated  because  of  the  decreasing  importance  of  the  money  held  by  nonreporting 
banks. 

2  "The  Problem  of  Monetary  and  Economic  Stability,"  Quarterly  Journal  of  Economics, 
February  1935;  see  chart  on  p.  192. 

•The  Berlin  Institute's  study  on  industrial  investment  and  financing  quoted  in  a 
previous  note  presents  a  chart  which  compares  the  increase  in  industrial  buildings  (Fabriken 
und  Werkstatten)  in  32  cities  from  1895  to  1918  with  pig-iron  consumption,  trend  being 
eliminated.  The  covariation  is  striking,  with  industrial  building  construction  slightly 


EXPENDITURE,  WAGES,  CUSTOMERS'  BALANCES      599 


meats  of  loans  and  discounts  is  not  wanting,  however,  especially  until 
the  end  of  the  second  Kondratieff,  although  it  reveals  itself  more  readily 
to  an  analysis  of  underlying  business  situations  than  it  would  to  formal 
measurement.  The  covariation  of  New  Security  Listings — intended  to 
indicate  roughly  fluctuation  of  households'  investment — with  loans  and 
discounts  is,  except  in  the  early  nineties,  fully  as  strong  as  we  can  expect 


LOANS  AND  DISCOUNTS 
NATIONAL  BANKS 
OUTSIDE  NEW  YORK 


INVESTMENTS  / 
NATIONAL  / 
BANKS  OUTSIDE 
NEW  YORK  / 


1870  75  I860  85  1890          95  1900          05  1910      13 

CHART  XXXIV.— United  States  (see  Appendix,  p.  1063). 

if  we  bear  in  mind  the  difference  in  the  "temperaments"  of  the  two  series 
and  the  difference  in  structure  of  the  "resonators"  of  which  they  are  the 
products.  Firms'  real  investment  is  again  represented  by  an  index  of 
the  production  of  industrial  equipment.  Its  graph  consistently  draws 
away  from  that  of  loans  and  discounts,  but  otherwise  varies  with  the 
latter,  though  again  in  the  manner  agreeable  to  its  temperament — a 
strong  jerk  in  the  one  often  corresponding  to  a  mere  kink  in  the  other. 
Both  series  display  the  two  Kondratieffs  in  much  the  same  way,  clearly 

but  almost  consistently  preceding.  Industrial  building  would  on  this  showing  turn  out 
to  be  an  even  better  index — or  forecaster — than  pig-iron  consumption, 


600  BUSINESS  CYCLES 

show  the  Juglars,  move  together  in  the  Kitchins.  The  reader  should 
satisfy  himself  that  such  exceptions  as  seem  to  occur  are  no  more  than 
understandable  lags.  Even  the  fact  known  to  ail  of  us  from  everyday 
experience,  that  there  is  such  a  thing  as  emergency  borrowing  and  that 
every  well-connected  concern  can  go  pretty  far  in  this  direction,  need 
hardly  ever  be  invoked. 

The  time  shape  of  loans,  therefore,  corresponds — in  the  long  and 
in  the  short  run ;  in  trend  and  in  all  the  cycles — as  perfectly  to  expectation 
from  our  model  as  the  time  shape  of  deposits  that  it  controls.  But 
again,  if  we  go  on  to  say  that  their  behavior  is  illustrated  and  explained 
by  the  behavior  of  the  pig-iron  consumption  or  equipment-production 
series,  we  are  adding  something  to  the  testimony  of  our  charts.  Repeat- 
ing the  corresponding  argument  submitted  in  the  case  of  system  expendi- 
ture, we  have  once  more  to  recognize  that,  although  perfectly  safe 
and  supported  by  common  opinion,  the  proposition  that  investment  is  in 
prosperity,  though  not  in  revival,  the  propeller  of  general  business  activity 
and  hence  controls  the  behavior  of  loans  and  balances,  is  such  an  addition 
from  other  evidence,  an  addition  moreover  which  is  not  so  sheltered 
by  triviality  as  the  proposition  that  loans  control  deposits  and  not 
deposits,  loans.1  Less  safe,  however,  is  the  further  addition  that  loans 
substantially  contribute  to  the  financing  of  real  investment,  and  still 
less  the  final  one  that  they  are  the — logically  primary — source  of  real 
investment  incident  to  innovation.  As  far  as  time-series  evidence  goes, 
it  would  be  just  as  possible  to  say  that  it  is  the  movement  of  loans  and 
deposits,  itself  determined  by  monetary  conditions  and  policies,  which 
induces  and  controls  real  investment,  or  that  initiative  is  not  with  entre- 
preneurs but  with  banks. 

This  issue  can  be  decided  only  by  theoretical  and  historical  analysis 
of  the  capitalist  process,  such  as  has  been  attempted  in  earlier  chapters, 
where  it  has  also  been  pointed  out  that  the  fact  that  loans  and  discounts 
mainly  finance  current  operations — in  each  cycle  perhaps  preponderately 
those  of  the  Secondary  Wave — and  even  consumers'  expenditure,  con- 
tradicts our  interpretation  in  appearance  only,  and  that  the  extent  to 
which  they  actually  serve  the  purpose  of  financing  long-time  commit- 
ments and  in  particular  industrial  plant  is  likely  to  be  underrated, 
because  that  purpose  hides  itself.  We  have  seen  by  such  outstanding 
examples  as  the  rise  of  the  automobile  industry  that  it  need  not  be  the 
entrepreneur  who  borrows  from  a  bank,  but  that  the  burden  can  be 
shifted  to  the  firms  who  furnish  the  materials  or  market  the  product; 
or  that  the  stock  exchange  speculator  or  the  investor  who  takes  new  issues 

1  Of  course,  in  another  sense  that  proposition  is  not  trivial,  viz.,  in  the  sense  that  it  is 
loans  and  not  member  banks'  investment  which  dominated  the  behavior  of  deposits  in  the 
period  under  discussion, 


EXPENDITUKE,    WAGES,    CUSTOMERS*    BALANCES          601 

may,  and  very  often  does,  borrow  instead  of  the  entrepreneur.  But 
we  have  also  seen  in  the  historical  chapters  plenty  of  indications  of  bank 
loans'  being  directly  used  for  the  purposes  of  innovation,  especially  in 
this  country,1  less  so  in  Germany,  least  of  all  in  England.  In  fact,  if  the 
reader  will  refer  to  the  list  of  the  methods  of  financing  system  expendi- 
ture at  the  beginning  of  this  section,  he  will  have  no  difficulty  in  satis- 
fying himself  that  it  must  be  so.  Though  borrowing  from  banks  is  not 
the  only  way  of — directly  or  indirectly — financing  innovation,  and 
though  bank  loans  also  serve  many  other  purposes,  our  schema  cannot  be 
so  very  far  from  practice  after  all;  for  most  of  those  methods  are  either 
inadequate  or  not  available  for  the  purpose  of  innovation  which  starts 
from  scratch. 

Glancing  over  our  list  of  methods  of  financing,  we  readily  see  that 
deflection,  overspending,  and  the  use  of  temporarily  available  funds 
will  neither  be  open  to  a  new  firm  nor,  in  general,  prove  sufficient  to 
finance  large-scale  plant  and  equipment.  Accumulation  and  saving 
provide  the  means  for  consolidating  rather  than  for  building  up  industrial 
ventures.  In  any  case,  they  owe  their  actual  importance  only  to  previous 
entrepreneurial  success  and  serve,  besides  repayment  of  debt,  induced 
expansion  rather  than  the  creation  of  entirely  new  things,  although  the 
latter  function  has  always  played  a  considerable  role  from  the  times  of 
primitive,  through  the  times  of  competitive,  to  the  times  of  "trustified'* 
capitalism.  The  same  applies  to  selling  assets  to  non-banks.  Neglecting 
items  of  secondary  importance,  we  are  thus  in  fact  left  with  selling 
assets  to,  and  borrowing  from,  banks,  as  influenced  by  borrowing  from, 
and  lending  to,  foreign  countries.  A  rough  correspondence  between 
innovation  and  loans,  still  more  between  innovation  and  deposits,  is 
hence  not  as  improbable  as  it  might  seem  at  first  sight. 

1  Some  further  indications  may  be  noticed,  however.  Mr.  Moulton,  bolder  than  the 
present  writer,  has  actually  presented  an  estimate  of  the  relative  importance  of  the  advances 
made  by  banks  in  1916,  that  went  to  "fixed  capital  operations,"  and  found  them  to  be  55  per 
cent  in  the  case  of  national,  62.1  per  cent  in  the  case  of  state  banks,  and  68.2  per  cent  in 
the  case  of  trust  companies  (Journal  of  Political  Economy,  June  1918).  The  responsibility 
is  Mr.  Moulton's.  Or  see  C.  O.  Hardy  and  Jacob  Viner,  Report  (to  the  Treasury  Depart- 
ments), on  the  Availability  of  Bank  Credit  in  the  Seventh  Federal  Reserve  District,  which 
states  that  many  short-term  loans  (not  to  speak  of  mortgage  loans,  of  which  those  on  non- 
farm  land  are  certainly  relevant  to  our  argument),  being  in  fact  renewed  indefinitely,  are 
used  by  borrowers  for  the  purpose  of  financing  expenditure  on  plant  and  equipment.  Part 
of  these  loans  must  finance  enterprise  in  our  sense. 


CHAPTER  XII 

The  Rate  of  Interest1 


A.  Earlier  Argument  Resumed. — Resuming  our  argument  about  the 
rate  of  interest  where  we  left  off  in  Chap.  Ill,  Sec.  E,  and  recalling  that 
we  look  upon  it,  on  the  one  hand,  as  a  premium  on  present  over  future 
means  of  payment  or  "balances"  and,  on  the  other  hand,  as  a  coefficient 
of  tension  in  the  system,  we  will  first  of  all  develop  some  of  the  aspects 
of  the  theory  there  presented.  It  is  convenient  to  take  the  first  step 
under  cover  of  all  the  simplifications  of  the  pure  model:  once  more  nobody 
is  to  borrow  except  entrepreneurs,  nobody  is  to  lend  except  member  banks 
which  create  the  required  "funds"  ad  hoc,  all  other  business — consisting 
only  of  purchases  of  producers'  goods  and  sales  of  consumers'  goods — 
being  financed  from  previous  receipts.  Excluding,  moreover,  for  the 
moment  what,  in  Chap.  Ill,  Sec.  E,  has  been  defined  as  the  Open  and  the 
Central  Markets,  we  have  at  least  the  comfort  of  facing  one  rate  of 
interest  only,  which  no  doubt  may  differ  as  wages  do  in  different  parts  of 
the  system,  but  can,  nevertheless,  be  used  in  the  same  way  as  the  wage 
rate.  While  all  these  simplifications  will  be  dropped  within  this  chapter, 
we  defer  to  the  next  the  consideration  of  international  capital  movements 
and  as  much  as  possible  also  of  everything  that  happens  within  the  walls 
of  member  banks'  committee  rooms  beyond  acceptance  or  refusal  of 
applications  submitted  by  entrepreneurs  on  their  own  initiative.  In 
other  words,  in  this  chapter  we  deal  primarily  with  closed  domains  and 
with  a  "passive"  banking  system,  in  order  to  see  how  great  an  expanse 
of  observed  fact  can  be  covered  within  these  restrictions. 

1.  In  the  absence  of  consumers'  time  preference,  innovation,  or  the 
expectation  of  profits  in  our  sense  of  the  term,  is  now  the  only  factor  to 
produce,  in  an  otherwise  stationary  and  undisturbed  world,  a  positive  pre- 
mium on  present  balances  or  to  make  present  dollars  worth  more  than  the 
same  amount  of  future  dollars.  Looking  at  any  amount  of  dollars  plus 
the  profit  an  entrepreneur  hopes  to  realize  by  means  of  it  as  bis  "demand 
price"  for  that  amount,  we  seem  to  be  able  to  build  a  curve  or  schedule 
of  entrepreneurial  demand  for  present  dollars,  the  ordinates  of  which  are 
1  Reference  should  be  made  again  to  Theory  of  Economic  Development,  Chap.  V.  But 
the  reader  is  once  more  reminded  that  many  of  our  propositions  are  not  dependent  upon 
the  writer's  theory  of  interest, 

602 


THE  RATE  OF  INTEREST  603 

expressed  in  terms  of  future  dollars,  and  to  conclude  in  the  familiar  way 
that  interest  will  equal  marginal  profit,  i.e.,  not  merely  the  profit  of  the 
marginal  entrepreneur  but,  since  we  may  postulate  that  every  entre- 
preneur can  continuously  vary  the  size  of  his  commitment,  the  marginal 
increment  of  prospective  profit  for  all.  But  such  a  construction  means 
very  little  in  our  case,  although  it  has  the  merit  of  showing  the  nature  of 
the  link  between  our  monetary  theory  of  interest  and  the  general  theory 
of  the  industrial  process.  We  have  not  only  to  recognize,  as  we  always 
have,  that  no  causal  significance  attaches  to  marginal  values  but  also 
that  all  the  significance  they  have  lies  in  the  setting  up  of  equilibrium 
conditions.  Our  case,  however,  is  essentially  one  of  disequilibrium  and 
the  circumstances  that  determine  marginal  profits  are  being  changed, 
and  known  to  be  changing,  every  moment  by  the  very  process  we  are 
trying  to  describe.1  What  we  can  hope  to  get  at,  therefore,  is  at  best  a 
rough  equivalence  between  interest  and  marginal  prospects  of  profits 
or  a  condition  not  of  equilibrium  in  the  full  and  proper  sense  of  the  term 
but  of  adapted  variation  within  our  process,  i.e.,  of  covariation  such  as  will 
not  by  itself  enforce  additional  adaptation  of  other  elements  of  the  system. 
This  is  all  we  have  to  offer  in  place  of  the  various  definitions  of  the  con- 
cept, Equilibrium  Rate  of  Interest.2  The  true  equilibrium  rate,  i.e., 
the  rate  that  would  obtain  in  a  stationary  process  perfectly  equilibrated 
in  all  its  elements  and  displaying  no  tension  at  all,  would  as  we  know  be 
zero.3 

Even  if  we  decide  to  neglect  these  difficulties  and  to  speak  of  such  a 
demand  schedule  after  all,  we  have  still  to  recognize  that,  more  than  any 
other,  it  shifts  violently  in  the  cycle.  This  trivial  consequence  of  any 

1  That  the  analogy  with  the  demand  schedule  for  a  factor  is  treacherous  is  also  obvious 
from  the  fact  that  well-behaved  demand  schedules  presuppose  invariant  production  func- 
tions.    In  our  case,  production  functions  are  being  altered  or  new  ones  are  being  built  up. 
Moreover,  capital  in  our  sense,  not  being  a  factor  of  production,  enters  neither  the  existing 
ones  nor  the  new  production  functions. 

2  As  pointed  out  in  Chap.  Ill,  Sec.  E,  that  condition  to  some  extent  stands  in  the  place 
of  the  two  conditions  that  are  required  in  the  Wicksellian  system,  viz.,  first,  that  the  natural 
rate  of  interest  should  be  in  equilibrium — for  example,  that  the  rate  of  time  preference 
should  equal  the  rate  of  technological  superiority  of  present  over  future  units  of  original 
means  of  production — and,  second,  that  the  monetary  rate  should  equal  that  natural  rate. 
For  a  development  (and  improvement)  of  this  theory  see  G.  Myrdal,  Der  Gleichgewichts- 
begriff  als  Instrument  der  geldtheoretischen  Analyse  (Beitrage  zur  Geldtheorie,  ed.  by 
F.  A.  v.  Hayek,  1933).     From  theories  of  the  equilibrium  rate  of  interest  we  should  distin- 
guish theories  of  the  equilibrating  rate  of  interest,  for  it  is  obvious  that  the  two  do  not 
necessarily  coincide.     A  rate  that  fulfills  the  ordinary  conditions  of  equilibrium,  for 
instance,  a  rate  that  clears  the  market  of  savings,  might,  nevertheless,  be  the  cause  of 
disturbances,  for  instance,  through  its  action  on  price  level.     Whoever  holds  that  it  would 
and  that  another  rate  would  not  had  better  state  separately  the  conditions  that  define  the 
two. 

8  But,  once  more,  the  author  does  not  here  insist  on  this  proposition. 


604  BUSINESS  CYCLES 

analysis  that  admits  any  relation  at  all  between  interest  and  profits  in 
our  sense  is  also  so  obvious  to  any  observer  of  business  life  that  it  should 
be  superfluous  to  state  it.  Yet  it  is  constantly  overlooked.  The 
exaggerated  ideas  that  many  economists  entertain — though  perhaps 
somewhat  less  so  now  than  they  did  10  years  ago — about  the  effectiveness 
of  a  reduction  of  the  rate  of  interest  in  inducing  investment  is  under- 
standable only  on  the  hypothesis  that  they  assume — besides  a  full  measure 
of  ceteris  paribus — that  the  demand  schedule  for  funds  is  substantially 
invariant,  at  least  for  periods  as  long  as  Kitchins  and  perhaps  as  long  as 
Juglars.  The  familiar  calculations  first  stressed  by  Wicksell  of  the 
tremendous  difference  it  makes  to  railroad  building  whether  interest  is 
1  per  cent  higher  or  lower,  on  the  one  hand,  and  the  frequent  recurrence 
of  complaints  about  the  failure  of  interest  to  fall  adequately  in  recession 
and  depression,  on  the  other,  which  are  so  curiously  at  variance  with  the 
fact  that  new  investment  and  even  replacement  is  primarily — we  have 
met  exceptions — associated  with  relatively  high  and  rising  money  rates, 
prove  better  than  anything  else  how  completely  most  economists  still 
think  in  terms  of  expansion  within  unchanging  production  functions  and 
how  reluctant  they  are  to  recognize  that  in  the  cyclical  process  shifts  and 
distortions  of  the  schedule  of  demand  for  balances  are  much  more  impor- 
tant than  movements  along  it.  There  are  situations  in  which  zero 
interest  would  entirely  fail  to  call  forth  any  additional  demand. 

2.  These  shifts  of  the  demand  schedule  for  balances  are,  of  course, 
intensified  by  the  operations  that  constitute  the  Secondary  Wave.  This 
phenomenon  we  have  already  met  with  in  the  case  of  wages,  but  it  asserts 
itself  much  more  directly  and  much  more  strongly  here.  There  is,  how- 
ever, another  kind  of  shift  which  still  more  seriously  interferes  with  the 
application  of  the  demand-supply  apparatus  whenever  there  is  full 
employment  of  resources.  Then  a  net  increase  in  balances  created  in 
response  to  business  demand1  enforces  a  rise  in  price  level,  induces 
expectations  of  further  rise,  and  for  both  reasons  further  increase  in 
demand  for  balances,  so  that  that  demand  schedule  tends  to  shift  also  in 
function  of  the  amount  of,  and  increase  in,  the  balances  themselves, 
and  in  consequence  completely  ceases  to  be  independent  of  "quantity  of 
commodity,"  as  a  demand  schedule  should  be.  If  entrepreneurs  really 
were  the  only  borrowers,  we  should  not  commit  any  great  error  by 
neglecting  this  effect.  For,  as  we  have  seen,  entrepreneurial  action  is  not 
typically  taken  on  an  expectation  of  a  rise  in  price  level  or  under  the 
influence  of  a  price  level  that  has  risen  already.  Moreover,  although 
entrepreneurs  may  not  be  aware  of  the  falling  result  trend  in  price  level, 
they  cannot  but  be  aware  of  the  fall  to  be  expected  in  the  prices  of  their 

1  It  will  be  observed  that  this  is  not  equivalent  to  saying  that  any  net  increase  in 
balances  created  will,  even  with  full  employment  of  resources,  necessarily  have  that  effect. 


THE  RATE  OF  INTEREST  605 

own  products,  which  they  themselves  intend  to  bring  about  in  order  to 
use  their  innovations  to  the  full  and/or  to  undersell  old  firms;  and  this 
will  in  general  suffice  to  prevent  them  from  acting  on  a  hypothesis  to  the 
effect  that  prices  will  go  on  rising  forever.  But  actually,  since  there  is  at 
any  time  in  prosperity  a  large  mass  of  windfall  gains  in  prospect,  the 
general  willingness  of  business  to  pay  interest  will,  both  because  of  the 
mechanical  effect  of  rising  prices  on  the  money  cost  of  doing  business  and 
because  of  extrapolating  anticipations,  increase — and  correspondingly 
fall  in  depression — by  much  more  than  it  otherwise  would.  The  rate  of 
interest  will  hence  no  longer  be  explainable  by  the  fundamental  factor 
alone,  not  only  because  there  are  other  borrowers,  but  also  because  many 
of  them  strongly  react  to  the  rate  of  change  of  prices. 

It  follows  that  what  we  have  termed  the  Adapted  Rate  must  now  be 
defined  with  reference  to  a  "demand  schedule"  for  balances  in  terms  of 
expected  windfalls,  behind  which  the  entrepreneurial  profits  that  "ignite" 
the  process  may  at  times  entirely  disappear.  This  rate  will  differ,  in 
prosperity  positively,  in  depression  negatively,  from  the  one  that  would 
be  appropriate  to  entrepreneurs'  profit  expectations  over  time,  and 
change  literally  every  hour.  But  it  will,  in  general,  also  differ  from  what 
we  may  call  the  equilibrating  rates,  i.e.,  the  rates  that  would  tend  to 
counteract  disturbance.  For  the  adapted  rate  will  not  do  that  as  long 
as  prosperity  is  in  full  swing  and  as  long  as  the  balances  that  are  being 
created  in  every  point  of  time  incessantly  change  the  conditions  under 
which  the  corresponding  loans  were  applied  for.1  This  fact  underlies  all 
arguments  for  an  active  policy  of  banker's  banks  in  prosperity  and  for 
"punitive"  rates.  In  a  competitive  banking  system  member  banks  will 
charge  the  adapted  rate  and  then  "catch  up"  only  when  windfalls, 
as  well  as  profits,  begin  to  fall  in  the  course  of  the  cyclical  process. 

This  being  one  of  the  causes  of  what  is  usually  referred  to  as  the  Lag 
in  Rate  of  Interest,  we  will  take  the  opportunity  of  disposing  of  it  at  once. 
Discarding  the  spurious  lag  resulting  from  the  failure  of  students  of  the 
cycle  to  distinguish  between  revivals  and  prosperities,  and  also  the 
frictional  lag  that  is  naturally  incident  to  a  relation  such  as  establishes 
itself  between  a  bank  and  its  customers  and  is  intensified  on  the  upgrade 
by  the  pressure  of  public  opinion2  and  on  the  downgrade  by  risks  and 
fears,  we  have,  besides  the  one  mentioned  in  the  preceding  paragraph, 
a  fundamental  reason  for  expecting  that  interest  will  tend  to  lag  behind 
other  cyclical  symptoms.  It  can  best  be  described  in  terms  of  our  pure 
model.  There  prosperity  starts  from  zero  loans,  so  that  the  lending 

1  This  may  be  called  a  cumulative  process  and  bears  some  resemblance  with  what  is 
known  as  the  Wicksellian  cumulative  process.     Possibly  we  are  visualizing  the  same  facts 
as  Wicksell.     But  the  two  arguments  themselves  have  little  in  common. 

2  Frictional  lag  is,  however,  almost  entirely  absent  from  open  market  rates,  which  in 
fact  are  the  most  active  ones  by  far. 


606  BUSINESS  CYCLES 

power  of  banks  is  entirely  unused,  while,  at  least  in  perfect  equilibrium 
of  perfect  competition,  all  the  factors  of  production  are  used  to  optimum 
points.  In  reality  this  is  not  so,  of  course,  but  our  abstraction  still 
serves  to  show  why  full  employment  does  not  imply  that  full  use  be  made 
of  the  existing  machine  for  the  manufacture  of  credit  and  why  this 
machine  will  in  general  be  underutilized  in  a  neighborhood  of  equilibrium. 
Now  in  a  world  conforming  to  the  pure  model,  interest  would  also  start 
from  zero  and  hence  be  one  of  the  first  elements  of  the  system  to  move. 
But  since  there  is  actually  in  every  neighborhood  of  equilibrium  a  positive 
rate  at  which  banks  will  in  general  be  glad  to  expand  their  loans,  that 
rate  will  not  rise  immediately. 

3.  Thus  the  demand  curve  for  balances  is,  even  with  greatly  simpli- 
fying assumptions,  an  instrument  of  doubtful  value.  But  the  value  of  the 
corresponding  supply  curve  is  more  doubtful  still.  Certain  difficulties 
would  remain,  even  if  we  could  interpret  it  as  the  supply  curve  of  savings, 
as  older  doctrine  did.  But  at  least  there  would  then  be  some  kind  of  cost 
curve  to  derive  it  from  and  hence  at  least  some  definite  meaning  to  it. 
Since,  however,  this  is  out  of  the  question,  and  since  it  would  similarly 
not  help  us  much  if  we  tried  to  build  a  bankers'  cost  curve  in  terms  of 
increasing  risks,  the  natural  thing  to  do  seems  to  be  to  drop  "supply  "  and 
to  fall  back  upon  "quantity  existing."1  But  this  too  would  be  entirely 
unrealistic,  because,  as  has  been  explained  in  Chap.  Ill,  Sec.  D,  there  is 
never  any  such  thing  as  a  definite  quantity  of  bank  accommodation  avail- 
able, not  even  in  a  perfectly  competitive  banking  system — if  every  bank 
moves  pari  passu  with  the  others — while  in  an  imperfectly  competitive 
one  the  individual  bank  has  a  large  space  to  maneuver  in.  Boundaries 
are  not  lacking,  but  we  have  seen  how  very  elastic  they  are.  Hence,  it  is 
necessary  to  recognize  an  element  of  indeterminateness  in  the  problem  of 
interest,  which  is  precisely  the  theoretical  reason  why  regulating  the 
money  market  differs  in  kind  from  regulating  any  commodity  market. 

We  may  now  insert  the  two  other  major  items  that  stand  alongside 
the  created  balances — savings,2  and  funds  that  are  temporarily  available, 
either  because  they  are  being  assembled  for  lumpy  acts  of  expenditure  or 
because  their  owners  have  decided  to  restrict  operations.  We  recall  our 
concept  of  Temporary  Investment  and  note  for  future  reference  that  it 
plays  a  role  in  the  theory  of  the  rate  in  the  open  market,  where  those 

1  It  should  be  obvious  without  further  explanation  that  the  sum  total  of  deposits  would 
not  measure  that  quantity  or  the  "supply  of  bank  credit"  any  more  than  bank  loans 
measure  the  "demand"  for  it. 

2  The  rate  of  savings  could  be  made  a  function  of  the  rate  of  interest  only  by  very  severe 
ceteris  paribus  assumptions.     Even  then  it  would,  as  everyone  now  admits,  not  be  a  simple 
or  a  monotonically  increasing  function.     But  this  is  no  reason  to  deny  the  existence  of  a 
functional  relation.     The  same  applies  to  the  inverse  relation  which,  in  Chap.  Ill,  Sec.  A, 
we  were  able  to  deal  with  so  simply  because  we  assumed  otherwise  stationary  conditions. 


THE  RATE  OF  INTEREST  607 

funds  join  forces  with  the  otherwise  unused  means  of  member 
banks. 

Similarly,  we  may  now  insert  the  other  major  items  of  total  borrowing. 
These  consist,  first,  in  the  amount  of  borrowing  that  proceeds  from  the 
Secondary  Wave  already  mentioned,  the  requirements  of  induced  and 
adaptive  expansion,  and  also  those  of  current  business,  since  as  a  matter 
of  fact  current  business,  too,  is  partly  transacted  on  a  credit  basis. 
And,  in  the  second  place,  there  is  the  vast  amount  of  consumers'  and 
quasi-consumers'  borrowing,  in  which  we  include  the  emergency  borrow- 
ing by  firms;  the  cases  (which  are  numerous,  especially  in  the  agrarian 
sector)  of  borrowing  which,  while  in  name  productive,  mainly  serves  pur- 
poses of  consumption;  and  finally,  the  borrowing  of  public  and  private 
households.  The  cyclical  behavior  of  some  of  these  items  is  doubtful. 
We  have,  for  instance,  seen  reasons  to  believe  (Chap.  XI,  Sec.  B)  that  at 
times  private  households  borrow  more,  both  in  the  form  of  installment 
contracts  or  open  accounts  and  in  the  form  of  loans,  in  prosperity  than 
they  do  in  any  not  too  prolonged  depression.  But  this  may  not  be  so 
true  of  prewar  times  and  Europe  as  it  is  of  postwar  times  and  the  United 
States.  Government  borrowing,  as  far  as  it  is  incident  to  war  and 
preparation  for  war,  is  erratic.  We  might,  however,  also  expect  to  see  a 
cyclical  component,  owing  to  the  fact  that  revenue  falls  off  in  deep 
depression,  even  though  depression  expenditure  did  not  in  our  epoch 
play  the  role  it  came  to  play  after  the  World  War.  This  expectation  is 
not  entirely  disappointed,  witness  the  English  case  in  the  last  years  of  the 
Melbourne-Russell  government.  But  in  the  times  of  "sound"  fiscal 
policies  such  cases  were  the  exception,  not  the  rule.  Goschen's  budgets 
more  than  balanced  in  the  midst  of  depression.  The  heavy  borrowing 
of  the  German  Empire  occurred  in  the  prosperity  phase  of  the  third 
Kondratieff. 

It  will  be  observed  that  some  of  the  remaining  items  tend  to  intensify, 
others  to  mitigate,  those  cyclical  variations  in  interest  rate  that  we  would 
expect  from  the  entrepreneurial  impulse  alone.  But  it  will  be  also 
observed  that  with  the  exception  of  the  requirements  for  induced  expan- 
sion or  adaptation,  which  will  tend  to  keep  up  the  rate  of  interest  in 
recession,  there  is  little  reason  to  expect  that  they  will  substantially  alter 
its  time  shape. 

4.  From  its  source  interest  spreads,  as  we  have  seen,  all  over  the  sys- 
tem. A  premium  on  present  dollars  in  any  sector  is  sufficient  to  enforce  a 
general  premium  in  all.  Thus  interest  intrudes  into  every  transaction, 
calculation,  and  valuation,  turns  time  into  a  cost  factor,  and  becomes 
that  subtle  and  omnipresent  entity  that  acts  on  and  reacts  to  everything 
and  is  so  difficult  to  trace  in  all  its  protean  forms.  Every  unit  of  currency 
or  deposits,  wherever  placed  or  circulating,  has,  in  order  to  stay  where 


608  BUSINESS  CYCLES 

it  is  or  to  go  on  in  its  circuit,  to  resist  a  pull  toward  the  money  market, 
which  at  the  margin  is  measured  by  the  rate  of  interest.  This  point  may 
be  illustrated  by  means  of  a  Wicksteedian  demand  curve  if,  for  the 
moment  and  for  the  purpose  of  illustration  only,  we  assume  that  there  is 
such  a  thing  as  a  definite  quantity  of  (cash  and)  balances  in  existence, 
including  the  amount,  also  assumed  to  be  given  at  any  point  of  time, 
which  banks  can  technically  create  beyond  what  they  actually  have 
created  at  that  given  point  of  time.  The  Wicksteedian  demand  curve 
refers  price  not  to  the  quantity  of  a  commodity  that  buyers  are  willing 
to  take  at  that  price  but  to  existing  quantity,  i.e.,  that  quantity  plus 
the  quantity  which  owners  keep  at  that  price.  Accordingly,  we  may  look 
upon  every  unit  of  actual  or  potential  balances  as  "  offered"  on  the  money 
market  and  taken  from  it  either  by  its  owner  or  someone  else.  The 
owner  must  then  be  thought  of  as  paying  interest  to  himself,  either  in 
the  form  of  some  element  of  return  if  he  uses  his  money  in  his  business, 
or  in  the  form  of  some  satisfaction  (equivalent  to  the  loss  of  interest 
involved)  if  he  does  not.1  But  apart  from  being  applicable  to  the  case  of 
perfect  competition  only,  this  schema  presupposes  a  string  of  assumptions 
that  are  entirely  inadmissible  in  the  case  of  money.  Although  it  may  be 
used  to  clarify  that  delicate  point  at  which  our  theory  proceeds  from  the 
analysis  of  the  logical  sources  of  interest  to  the  analysis  of  interest  as  an 
all-pervading  phenomenon,  it  should  not  be  taken  too  seriously. 

In  this  sense  interest  may  indeed  be  said  to  hold  a  central  position 
in  the  system.  But  again,  precisely  because  it  does,  it  is  easy  to  slide 
into  exaggerations  of  the  influence  unilaterally  exerted  by  it  and  to  forget 
that  a  central  position  does  not  imply  a  key  position,  particularly  in  the 
cyclical  process  of  evolution.  Just  now,  however,  two  cognate  errors 
call  for  our  attention.2  With  the  growth  of  capitalist  mentality,  an 
obviously  useful  habit  has  developed,  the  beginnings  of  which  are  in  Ger- 
many, for  instance,  observable  since  the  fourteenth  century,  of  expressing 
any  return,  except  returns  to  personal  services,  as  a  percentage  of  a 
capital  value.  But  it  is  an  error  to  conclude  that  these  returns  are 
turned  into  interest  thereby  or,  worse  still,  that  they  are  the  basic  fact 
about  interest  and,  barring  temporary  discrepancies,  fundamentally 
determine  the  money  rate.  What  any  durable  product  yields — no  matter 
whether  a  consumers'  or  a  producers'  good,  though  rational  calculation 
is  understandably  more  in  evidence  with  respect  to  the  latter — once  it  has 

1  Viewed  from  this  standpoint,  interest  could  also  be  defined  as  the  price  that  it  is 
necessary  to  pay  to  a  holder  of  actual  or  potential  balances  in  order  to  induce  him  to  part 
with  "his  money."     Thus  we  meet,  for  a  moment  and  under  very  restrictive  assumptions, 
the  concept  of  interest  which  has,  in  the  General  Theory  of  Employment,  Interest  and 
Money,  been  adopted  by  Mr.  Keynes.     But  the  point  of  tangency  between  our  argument 
and  his  is  not  more  obvious  than  the 'divergence  of  the  curves. 

2  For  a  full  discussion  of  what  follows  see  Theory  of  Economic  Development,  Chap.  V. 


THE  RATE  OF  INTEREST 

come  into  existence,  is  and  remains  (temporary)  quasi-rent, 
may  express  it.  It  is  neither  the  same  thing  as  interest  nor  i 
equal  to  interest.  The  other  error  consists  in  the  interpretation 
discounting  process  by  which  the  capital  value  of  a  durable  good  is  arrived 
at  and  which  indeed  establishes  a  relation  of  mutual  dependence — ideally, 
equality  at  the  margin — between  quasi-rents  and  interest.  But  it  does 
so  by  applying  a  logically  prior  monetary  rate  of  interest — that  monetary 
rate  which  is  expected  to  prevail  during  the  life  of  the  good1  in  the  sector 
within  the  horizon  of  the  firm  or  household  concerned — to  the  sequence 
of  the  expected  installments  of  quasi-rent  evaluated  in  money,  and  not  by 
virtue  of  an  interest  logically  independent  of  money.  It,  hence,  pre- 
supposes the  monetary  rate,  instead  of  controlling  it. 

Of  all  the  ramifications  of  this  principle  of  evaluation  and  all  the 
qualifications  that  it  calls  for,  only  two  need  to  be  touched  upon.  First, 
we  have  again,  as  in  the  case  of  the  demand  schedule  for  balances,  to 
bear  in  mind  that  the  demand  schedule  for  durable  goods — -which,  though 
doubtful  enough,  is  nevertheless  more  like  what  a  demand  schedule  should 
be — is  subject  to  very  strong  shifts  and  distortions,  which,  in  the  course 
of  cyclical  phases,  are  more  important  than  any  movements  along  it. 
Professor  Moore's  rising-demand  curves  for  steel  and  other  products  that 
enter  into  durable  goods  (Chap.  X,  Sec.  A.)  give  a  good  idea  of  how  com- 
pletely the  latter  are  overshadowed  by  the  former.  This,  of  course,  is 
only  what  we  should  expect.  But  it  practically  implies  an  analogous 
proposition  about  the  influence  of  variations  of  interest.  Although  this 
does  not  impair  the  logical  validity  of  the  theorem  that  values  of  such 
goods  will,  ceteris  paribus,  rise  and  their  production  become  more  profit- 
able if  a  fall  in  the  rate  of  interest  occurs  or  is  anticipated,  it  does  impair 
its  relevance  to  our  subject.  The  outstanding  fact  that  is  really  relevant 
to  our  subject  is  the  positive  association  of  rate  of  interest  and  value  of 
capital  goods.  Our  motive  for  insisting  on  it  is,  again,  that  many  state- 
ments in  older  as  well  as  more  recent  literature  are  understandable  only  on 
the  hypothesis  that  it  has  been  overlooked.2 

1  That  life  is  not  a  (technological)  datum  but  a  variable  of  the  problem;  but  this,  like 
other  minor  points,  need  not  concern  us  here. 

2  The  objection  which  might  be  raised  against  the  above  argument,  viz.,  that  it  is  not  the 
money  or  the  natural  rate  that  counts,  but  the  difference  between  the  two,  misses  the  point, 
or,  if  natural  rate  is  but  another  name  for  profits  in  our  sense,  merely  reformulates  it. 
But  the  above  argument  should  not  be  interpreted  to  mean  that  financial  editors  are  wrong 
in  hailing  a  falling  rate  of  interest  as  a  hopeful  symptom  in  a  difficult  situation.     Although 
its  causal  importance  may  be  next  to  nothing — we  do  not  evaluate  it  at  zero,  however — 
it  is  a  symptom  of  decrease  of  strain,  of  progressing  adaptation  and  liquidation,  and  in 
many  cases  it  is,  and  especially  was  in  the  nineteenth  century  when  panics  were  allowed  to 
have  the  effect  of  forcing  up  the  rate,  a  sign  of  the  worst  being  over.     If  the  businessman 
clamors  for  cheap  money,  that  means  as  much  as  and  not  more  than  his  clamoring  for  cheap 
labor. 


610  BUSINESS  CYCLES 

A  second  point  turns  on  the  fact  that,  even  where  it  is  possible  to 
identify  the  rate1  at  which  it  is  rational  for  a  firm  or  a  household  to  dis- 
count a  given  series  of  quasi-rents,  we  in  general  find  a  considerable  dis- 
discrepancy  between  the  result  and  actual  values  of  durable  goods  or 
going  concerns.  There  are  many  obvious  reasons,  both  sociological  and 
other,  which  need  not  detain  us,  why,  e.g.,  rural  real  estate  at  times  sells 
much  above  the  price  our  principle  would  indicate.  But  one  of  these 
reasons  is  especially  important  for  the  subject  in  hand*  viz.,  the  risks  and 
chances  of  depreciation  and  appreciation  incident  to  ownership  of  an 
individual  piece  of  property.  Whoever  acquires  a  durable  good  commits 
himself.  He  chooses  one  out  of  an  innumerable  host  of  possible  sets  of 
risks  and  chances,  which  differ  not  only  as  to  the  objects  or  groups  of 
objects  to  be  bought,  but  also  as  to  the  point  of  time  at  which  to  effect 
the  purchase.  In  doing  so  he,  to  some  extent,  foregoes  the  opportunity 
of  better  sets  offering  themselves,  to  which,  after  having  taken  action, 
he  cannot  shift  over  at  all  or  to  which  he  can  shift  over  only  with  difficulty 
and  at  a  loss.  This  loss  is  peculiar  in  that  the  danger  of  it  is  present  even 
in  otherwise  perfectly  safe  investments.  For  this  the  quasi-rent  of  an 
eligible  set  must  provide  compensation  that  will  largely  account  for  the 
discrepancies  under  discussion.  Making  up  one's  mind,  therefore, 
involves  forming,  consciously  and  subconsciously,  a  great  many  anticipa- 
tions about  the  future  behavior  of  relevant  factors,  the  net  marginal 
result  of  which,  currently  corrected,  is  for  the  public  as  a  whole  theo- 
retically measured  by  those  deviations.2  One  of  these  anticipations  refers 
to  the  future  behavior  of  interest.  Inasmuch  as  this  may  be  the  domi- 
nating consideration,  we  can  perhaps  say  by  way  of  a  first  approximation 
that  deviations  of  the  relation  between  values  of  investments,  similar  in 
other  respects  but  involving  comitments  of  different  duration,  from  the 
relation  supplied  by  our  principle  of  discounting3  express  the  public's 
marginal  opinion  about  the  future  behavior  of  the  rate  of  interest. 

1  For  reasons  which,  if  not  obvious,  will  presently  appear,  that  is  often  difficult  and 
sometimes  hopeless.  Moreover,  in  highly  imperfect  situations  in  which  all  or  some  con- 
cerns, instead  of  behaving  like  drops  in  a  sea,  are  more  like  discrete  islands  conditions 
between  which  are  never  fully  equalized  and  which  have  their  special  markets  not  only 
for  their  products  but  also  for  their  factors,  it  is  quite  possible  that  some  of  those  concerns 
also  have  semidetached  money  markets  of  their  own,  partly  fed  by  their  own  accumula- 
tions or  the  savings  of  people — officers,  workmen — connected  with  them.  In  such  cases 
there  may  be  point  in  speaking  of  an  "internal"  rate  of  interest  as  distinct  from  either  the 
general  one  or  that  of  other  concerns.  Fundamentally,  however,  and  particularly  in  perfect 
conditions,  the  rate  to  discount  with  is  always  the  market  rate. 

8  As  we  have  seen  in  Chaps.  II  and  IV,  these  anticipations  may  act  as  disequilibrating 
factors,  and  this  as  much  if  they  are  borne  out  by  events  as  if  they  fail  to  come  true.  But 
this  is  not  necessarily  so,  and  the  case  in  which  they  have  an  equilibrating  effect  is  practi- 
cally more  important  than  the  others. 

3  It  is  easy  to  see  that,  even  if  valuation  of  durable  instruments  conformed  exactly  to 
that  rule,  the  values  of  instruments  of  unequal  durability  would  be  unequally  affected  by 


THE  RATE  OF  INTEREST  611 

5.  Now,  the  future  balances,  which  according  to  our  theory  are  the 
price  of  present  balances,  do  not  as  such  exist,  and  all  the  seller  of  present 
balances  for  the  time  being  receives  is  promises.  These  must  be  embodied 
in  legal  instruments,  the  types  of  which  vary  as  much  as  the  types, 
situations,  and  particular  purposes  of  borrowers  and  lenders — from  the 
savings  and  (as  far  as  it  is  not  simply  the  same  thing  as  a  demand  deposit) 
time  deposit,  the  note  and  bill,  the  open  account,  and  so  on  to  the  mort- 
gage, the  bond  and,  as  from  our  standpoint  we  have  to  add,  the  share. 
And  since  all  these  embodiments  of  future  balances  bear  from  the  actuarial 
and  the  investor's  standpoint  a  surface  resemblance  to  other  marketable 
sources  of  returns,  much  of  what  has  just  been  said  could  be  repeated 
for  them.  Actually,  the  same  operation  is  being  performed  on  a  sequence 
of  expected  interest  and  capital  payments — or  of  expected  dividend  pay- 
ments plus  expected  remainders  of  equity  values — as  on  a  sequence  of 
expected  quasi-rent  installments.  From  this  results  the  difference 
between  bond  rate  and  bond  yield,  for  example,  as  well  as  the  relation 
between  them.1 

The  fact  that  titles  to  future  balances  differ  so  much  in  economic 
significance  and  that  even  a  given  economic  configuration  may  be  formu- 
lated, and  the  resulting  claim  safeguarded,  in  so  many  different  ways, 
gives  rise  to  what  it  is  usual  to  call  the  Structure  of  Rates.  There  is  no 
objection  against  thus  emphasizing  that  fact.2  But  it  should  not  blind 
us  to  the  truth  that,  as  far  as  this  goes,  the  differences  between  the  con- 
stituent individual  rates  are  still  only  the  same  kind  of  thing  as  the  differ- 
ences between  prices  of  different  qualities  of  a  good  or  the  differences 

a  change  in  the  rate  of  interest.  Such  a  change,  therefore,  tends  to  alter  not  only  the  dis- 
tribution of  productive  resources  between  transient  and  durable  goods  but  also  the  time 
structure  of  the  latter  taken  by  themselves. 

1  The  reader  will  observe  that  from  that  perfectly  general  standpoint  it  is  no  longer 
possible  to  say  that  quasi-rents  will  be  discounted  in  such  a  way  that  the  net  return  result- 
ing from  this  process  will  be  marginally  equal  to  pure  interest  plus  risk  and  chance,  because 
for  that  purely  formal  purpose  the  premium  present  balances  can  command  bears  no  other 
logical  relation  to  their  "capital  value"  than  the  one  which  quasi-rents  (though  temporary; 
this  difference  disappears  in  the  discounting  operation)  bear  to  theirs.     What  ought  to  be 
said  now  is  that  net  return  of  whatever  kind — whether  interest  or  quasi-rent — tends 
toward  equality  at  the  margin  with  risk  and  chance  (with  the  addition,  perhaps,  of  such 
costs  as  the  act  of  purchase  involves,  such  as  taxes  or  commissions).     But  this  is  only  one 
of  the  many  relations  between  interest  and  other  quantities,  which  we  may  formulate  as 
soon  as  interest  pervades  a  system,  and  should  not  be  included  in  any  fundamental  explana- 
tion of  its  nature.     A  similar  formula  equally  true  in  a  sense  is  this:  We  have  seen  that 
interest  is  a  premium  of  present  over  future  balances  and  not,  as  Boehm-Bawerk  has  it,  of 
present  over  future  goods;  but  insofar  as  the  present  balances  are  intended  by  "borrowers" 
to  be  spent  on  present  goods,  it  might  be  said  that  interest  is  a  premium  of  present  balances 
over  present  goods.     However,  this  again  is  more  misleading  than  helpful. 

2  There  is,  however,  considerable  objection  to  replacing  the  rate  of  older  doctrine  by  an 
index  of  rates,  which  would  obliterate  all  that  is  most  significant  in  that  structure. 


612  BUSINESS  CYCLES 

between  wage  rates  within  an  economic  domain.  If  there  were  no  others, 
it  would  still  be  admissible  and  for  certain  purposes  necessary  to  speak 
of  one  rate  of  pure  interest  from  which  the  various  observable  ones  differ 
by  virtue  of  the  element  of  risk  and  chance  in  the  above  sense;  this,  in  fact, 
we  have  no  difficulty  in  doing  in  everyday  life. 

There  is,  however,  a  cleavage  that  is  peculiar  to  interest  and  that  goes 
deeper  than  that,  viz.,  the  cleavage  between  the  rates  in  the  three  markets 
mentioned  in  Chap.  Ill,  Sec.  E — the  Money  Markat  proper,1  the  Open 
Market,  and  the  Central  Market.  It  is  this  which  "structures"  rates 
fundamentally.  That  the  money  market  and  the  central  market  do  not 
stand  on  a  par  and  are  completing  to  each  other  and  not  competing  is 
obvious.  The  open  market  owes  its  distinctive  characteristic  to  the 
nature  of  the  funds  that  flow  into  it  and  that  are  being  lent,  although  at 
the  same  time  bound  to  other  purposes.  The  cleavage  between  this  and 
the  money  market  is  neither  in  fact  nor  in  logic  as  sharp  as  the  one 
between  the  latter  and  the  central  market.  Communication  is  much 
more  direct,  and  money  and  open  markets  compete  with  each  other. 
The  distinction,  however,  is  so  important  and  emphasizes  so  essential 
an  element  of  the  modern  financial  mechanism  that  we  cannot  afford  to 
miss  it. 

But  by  thus  recognizing  the  existence  in  fact  as  well  as  in  theory  of  a 
fundamental  division  into  three  markets  of  the  total  volume  of  trading  in 
balances,  we  do  not  recede  from  our  protest  lodged  in  the  last  paragraph 
of  Chap.  Ill,  Sec.  E  against  another  attempt  at  drawing  such  a  funda- 
mental line,  viz.,  between  the  money  and  the  capital  market.  We  can, 
of  course,  distinguish  short-  and  long-term  financing  and,  for  purposes 
of  description,  can  think  of  the  latter  as  concentrated  in  a  special  market, 
which  we  may  call  as  we  please — capital  market  or  stock  exchange,2  for 
instance.  This  special  market  or  rather  its  component  parts  are,  more- 

1  The  writer  has  perhaps  to  apologize  for  his  clumsy  terminology,  needlessly  departing 
from  that  of  financial  practice,  which  uses  the  term  money  market  synonymously  with  open 
market.     It  is  more  important  that  the  above  tripartite  division  does  not  seem  to  leave 
room  for  lending  by  other  agencies  than  banks  and  quasi-banks.     The  lending  transactions 
of  households — there  is  and  has  been,  for  instance,  in  Germany  a  nonnegligible  amount 
of  mortgage  loans  given  directly  by  households  to  other  households  or  firms — and  those 
transactions  of  firms  which  really — and  not  only  apparently,  i.e.,  by  means  of  bank  credit 
extended  to  the  lending  firms — finance  open  accounts,  installments  and  so  on,  should  be 
considered  as  an  appendage  to  the  money  market  proper,  although,  of  course,  they  produce 
different  effects.     In  many  cases  firms,  especially  wholesale  firms  ("merchants"),  must  be 
looked  upon  as  quasi-banks:  in  fact  they  often  develop  into  private  banks.     We  can 
thus  peremptorily  dispose  of  this  matter  because  it  does  not  greatly  affect  the  drift  of  our 
argument. 

2  The  stock  exchange,  which  will  not  come  in  for  discussion  until  the  end  of  Chap.  XIII, 
is  for  us  simply  an  institution  for  the  handling  of  transactions  that  belong  partly  to  the 
money  market  and  partly  to  the  open  market. 


THE  RATE  OF  INTEREST  613 

over,  obvious  realities  that  raise  problems  peculiar  to  themselves,  and, 
in  particular,  produce  their  own  gross  rates  of  interest  in  the  sense  that 
is  ordinarily  meant  when  one  speaks  of  the  structure  of  rates.  In  this 
sense  we  shall  presently  use  this  distinction,  although  in  doing  so  we  shall 
merely  speak  of  different  sectors  of  the  money  market ;  but  no  significance 
deeper  than  that  attaches  to  it.  To  say,  for  instance,  that,  while  balances 
are  the  commodity  in  the  money  market,  something  that  is  not  balances 
constitutes  the  commodity  in  the  capital  market  (savings  or  capital  goods) 
is  not  only  utterly  unrealistic  but  wrong. 

Even  if  it  be  not  held  that  the  money  and  the  capital  market,  in  the 
sense  of  the  theories  just  glanced  at,  differ  in  fundamentals,  and  if  it 
be  correctly  recognized  that  in  both  it  is  balances  and  nothing  else  that 
is  being  traded  in,  and  that  such  difference  as  there  is  consists  only  in 
the  various  accessories  of  the  garb  in  which  the  promises  of  future  pay- 
ments put  in  their  appearance,  it  is  dangerous  and  often  indicative  of 
faulty  analysis  to  overstress  that  difference.  This  easily  suggests  that 
money-market  transactions — in  this  sense — have  little  to  do  with  long- 
term  financing  and  that  capital-market  transactions  never  serve  the  pur- 
poses of  current  business  or  that  the  former  have  only  to  do  with  the 
effecting  of  payments  (Zahlungskredit)  and  the  latter  only  with  real 
investment.  It  need  not  be  emphasized  again  how  wrong  all  that  is. 
Moreover,  looking  not  at  purposes  but  merely  at  the  instruments  in  which 
future  balances  are  embodied,  it  is  not  less  wrong  to  think  that  long 
maturities  must  necessarily  be  financed  by  long-term  funds.  On  the  con- 
trary, it  is  one  of  the  most  characteristic  features  of  the  financial  side  of 
capitalist  evolution  so  to  "mobilize"  all,  even  the  longest,  maturities  as 
to  make  any  commitment  to  a  promise  of  future  balances  amenable  to 
being  in  turn  financed  by  any  sort  of  funds  and  especially  by  funds  avail- 
able for  short  time,  even  overnight,  only.  This  is  not  mere  technique. 
This  is  part  of  the  core  of  the  capitalist  process. 

The  capitalist  process  develops,  along  with  the  money  market  (in  our 
sense),  perfect  negotiability  of  all  instruments  of  credit,  whatever  their 
legal  form  may  be.  Already  (Chap.  VI,  Sec.  A)  we  have  insisted  on  the 
importance  of  the  evolution  in  the  course  of  the  fifteenth  and  sixteenth 
centuries  of  the  negotiable  note  or  bill.  But  long-term  financial  contracts 
such  as  are  embodied  in  a  bond,  a  mortgage,  and  also  from  our  standpoint 
in  a  share,  became  later  just  as  negotiable  in  principle.  In  fact,  we  can 
say,  if  we  except  a  small  minority  of  cases,  that  for  the  individual  "seller" 
of  balances  no  investment  in  any  such  title  is  fixed  in  the  sense  that  it 
binds  his  money  for  a  protracted  or  any  definite  time.  Fixity  is  thus 
reduced  to,  or  rather  replaced  by,  the  cost  incident  to  actual  mobilization 
and  the  risk  of  not  being  able  to  "buy  back"  the  balance  without  loss. 
Of  course,  a  new  bond  or  share  may  not  attain  perfect  salability  until  it 


614  BUSINESS  CYCLES 

has  acquired  standing;  its  special  market  may  remain  very  limited  forever; 
and  for  these  and  other  reasons  its  purchase  may  involve  the  decision  to 
keep  it  for  a  considerable  time  or  not  to  sell  it  at  all  except  when  an 
occasion  offers  itself.  But  this  does  not  alter  the  principle  that  in 
capitalist  society — feudal  society,  for  instance,  behaved  differently — all 
equities  and  claims  can  normally  be  sold  at  will,  irrespective  of  the  purposes 
that  the  sums  originally  paid  in  may  serve — irrespectively  also  of  the  pur- 
poses any  goods  serve  that  may  correspond  to  them — hence  bought  by 
means  of  short-term  funds.  Bonds,  for  instance,  thus  become  a  vehicle 
of  the  shifting  of  balances,  which  only  technically  and  by  degree  differs 
from  short-term  instruments.  As  soon  as  this  is  realized,  doubts  arise 
about  the  very  existence  of  a  distinct  thing  to  be  called  the  long-term 
interest  rate.  The  contractual  rate  on  long-time  instruments  has,  of 
course,  a  certain  right  to  that  name.  But  at  the  moment  a  contract  is 
concluded  this  rate  is  simply  a  function  of  all  the  conditions  of  the  money 
market — barring  risks  and  chances  incident  to  financing  the  particular 
proposition — which  are  expressed  by  the  short  rates.  And  as  soon  as  the 
bonds  have  come  into  existence,  it  is  their  yield,  again  a  function  of  short 
rates,  that  counts  and  not  the  contractual  interest.  Braving  some  danger 
of  misunderstanding,  we  may  hence  go  so  far  as  to  say  that  there  exists 
no  such  thing  as  the  long-term  rate  and  that,  if  we  nevertheless  wish  to  use 
the  concept,  the  thing  we  ought  to  mean  is  some  kind  of  "trend  value"1 
of  short  rates. 

B.  Discussion  of  Various  Rates. — Information  about  the  historical 
course  of  interest  exists  from  very  early  times,  even  for  the  theocracies  of 
antiquity.  About  the  Graeco-Roman  world  we  know  much  more.  We 
know,  for  instance,  what  Roman  societates  publicanorum  did  to  unfor- 
tunate provinciates,  and  have  plenty  of  material  about  both  legislation 
and  practice.  The  point  that  strikes  us  first  of  all  is  the  preponderance  of 
lending  for  purposes  of  consumption.  Whether  protected  or  enslaved, 
the  debtor  is  typically  either  an  aristocratic  scapegrace2  or  else  a  poor  man 
submerged  in  misfortune.  It  is  this  aspect  that  legislators  and  jurists 
primarily  thought  of  and  that  they  transmitted  to  the  Christian  doctors 
and  canonists — with  all  the  more  success  because  these  found  exactly 
the  same  aspect  exclusively  stressed  by  the  Philosophus,  i.e.,  Aristotle. 
In  other  words,  they  and  their  medieval  successors  thought  primarily  of 

1  Lest  this  should  seem  to  contradict  a  later  statement,  it  is  as  well  to  point  out  that  that 
term  is  here  used  in  a  nontechnical  sense. 

2  In  Rome,  two  distinct  types  are  clearly  discernible:  politicians,  whose  career  greatly 
depended  on  lavish  expenditure  on  drcenses  and  other  things,  G.  Julius  Caesar  being  the 
outstanding  example  for  the  huge  element  of  risk  that  entered  into  lending  to  this  type; 
and  the  young  man  of  fashion,  whose  doings  frightened  the  old  gentlemen  of  the  senate 
into  passing  the  senatus  consultum  Macedonianum  in  order  to  avoid  being  killed  by  sons  in 
desperate  straits. 


THE  RATE  OF  INTEREST  615 

usury,  and  our  theory  lends,  in  fact,  some  support  to  their  arguments. 
The  discovery  that  debt  may  open  up  the  road  to  wealth  was  not,  how- 
ever, deferred  until  modern  times.  The  outstanding  case  of  recognition 
of  this  truth  is  afforded  by  a  contract  that  the  Romans  took  from  the 
Greeks,  the  foenus  nauticum.  This  provided  a  method  for  the  financing 
of  maritime  trade  mainly  by  removing  or  relaxing  the  restrictions  on  the 
rate  of  interest  in  consideration  of  the  clause  that  the  "entrepreneur's" 
obligation  as  to  both  interest  and  capital  lapsed  in  case  his  venture  failed 
to  succeed,  i.e.y  that  he  owed  only  if  ship  and  cargo  landed  safely.  If  it  be 
recalled  that  we  allocate  risk  to  the  capitalist,  it  will  be  seen  how  perfectly 
this  contract  expresses  the  aspect  that  to  us  is  the  essential  one. 

Data  become  plentiful  from  the  fourteenth  century  on,  but  material 
sufficient  to  derive  time  series  is  not  available,  so  far  as  the  writer  knows, 
before  the  eighteenth  century,  during  which  we  can  follow  the  prices  of 
the  English  consols.  Except  for  this,  serviceable  series  mostly  start  in 
the  second  quarter  of  the  nineteenth  century,  although  in  important 
sections  not  before  the  second  half.1 

1.  Success  of  factual  investigation  is  in  this  field  much  interfered  with 
also  by  circumstances  other  than  mere  insufficiency  of  material.  First, 
even  if  we  disregard  the  problem  of  the  influence  upon  a  given  rate  of 
interest  which  expectation  of  a  rise  or  fall  in  the  price  level  or  in  the  rate 
of  interest  itself  may  have,  it  is  impossible  to  isolate  the  element  of  risk 
and  chance.  The  high  rates  which  ruled  in  the  Middle  Ages,  for  instance, 
are  largely  attributable  to  the  fact  that,  particularly  in  loans  to  princes, 
the  lender  had  to  consider  the  likelihood  that  he  would  not  be  repaid  at  all. 
It  is  impossible  to  tell  how  much  would  have  remained  of  the  80  per  cent 
that  Frederick  the  Fair,  the  gallant  king  of  Germany  and  duke  of  Austria 
(fourteenth  century),  is  reported  to  have  "paid,"  had  the  financial  habits 
of  that  prince  been  more  regular.2  Cases  of  this  kind  abound  at  all  times, 

1  See,  however,  the  series  beginning  September  1795  of  prices  of  60-day  bills  on  London 
at  Boston  and  New  York  in  Smith  and  Cole,  Fluctuations,  App.  E.     Their  discount  series 
(ibid.,  Table  74)  starts  in  1831.     The  best  (monthly)  American  series,  the  interest  rate  on 
60  to  90-day  commercial  paper  in  New  York,  is  in  the  writer's  opinion  the  best  in  any 
country.     It  has  been  used  by  Professors  Crum  and  Frickey,  and  the  writer's  work  has  been 
primarily  based  on  it.     For  Germany,  see  Kahn,  Die  Geschichte  des  Zinsf usses  in  Deutsch- 
land  seit  1815  (1894);  Homburger,  Die  Entwicklung  des  Zinsf  usses  in  Deutschland,  1870- 
1903  (1905);  Albrecht,  Die  Geschichtliche  Entwicklung  des  Zinsfusses  in  Deutschland, 
1895-1908  (1910);  and  finally,  Voye,  Ueber  die  Hohe  der  verschiedenen  Zinsarten  (1902), 
who  for  Prussia  covers  the  period  from  1807  to  1900.  Together,  these  authors  tell  a  pretty 
complete  story. 

2  This,  by  the  way,  should  be  borne  in  mind  by  historians  who  report  on  the  sorry  facts 
of  usury.     To  the  interest  rates  recorded  correspond  the  records  of  the  failures  of  financial 
houses  that  did  that  sort  of  business  and  were  in  many,  if  not  most,  cases  bankrupted  by  it. 
On  the  other  hand,  we  have  seen  that  lenders  reaped  other  advantages  besides  interest 
(Chap,  VI). 


616  BUSINESS  CYCLES 

and  so  do  cases  in  which  the  lender,  by  hiding  the  actual  interest  behind 
some  supplementary  charge  or  concession,  avoids  quoting  a  figure  that 
would  shock  public  opinion.  All  this  has  an  important  consequence  for 
the  behavior  of  some  of  our  series,  which,  moreover,  change  their  char- 
acter as  time  goes  on.  Ever  since  the  English  government,  for  instance, 
began  to  enjoy  that  confidence  which  it  justified  for  so  long  a  time,  the 
price  of  consols  varied  much  as  we  should  expect  in  a  safe — although  in 
our  sense  never  riskless1 — annuity.  But  this  was  not  so  for  the  greater 
part  of  the  eighteenth  century,  when  we  find  consols  varying,  though  not 
so  strongly,  in  the  same  direction  as  stocks  did.  Similarly,  even  the 
highest  class  of  American  railroad  and  industrial  bonds  display  the  influ- 
ence, in  sections  of  our  period,  of  a  decreasing  element  of  risk. 

Moreover,  regional  differences  are  sometimes  so  important  as  to  make 
it  difficult  to  speak,  even  for  one  and  the  same  type  of  loan,  of  the  rate 
of  interest  in  a  country,  while  interregional  and  international  capital 
movements  interfere  with  the  behavior  of  interest  rates  in  any  particular 
country.  Again,  taxation,  fear  of  taxation,  and  the  general  attitude  of 
public  administration  toward  interest  will  all  have  effects  which  are  not 
easy  to  discern,  because  they  may  assert  themselves  with  lags  and  in 
very  many  forms,  differently  also  for  different  kinds  of  titles.  In  cases  in 
which  placing  of  bonds  is  difficult  and  costly,  or  where  intervention  on 
the  market  is  provided  for,  there  is  an  element  of  remuneration  for  these 
services.  Some  rates  are  very  slow  in  responding  to  changes  of  the  situa- 
tion. Others  mean  different  things  or  have  varying  importance  not  only 
in  different  countries  but  also  in  the  same  country  at  different  epochs. 
The  American  commercial  paper  rate  is  a  conspicuous  example. 

Even  if  bankers'  banks  did  no  business  except  with  member  banks, 
it  would  not  follow  that  there  is  only  one  rate  in  the  Central  Market. 
A  central  bank  is  a  discriminating  monopolist,  although  not  a  monopolist 
that  aims  at  maximizing  net  revenue.  For  many  purposes  this  can  be 
neglected,  however.  Similarly,  the  Open  Market  evolves  several  rates 
which  at  times  differ  significantly  but  still  present,  normally  at  least, 
a  picture  so  uniform  as  to  warrant  speaking  of  the  open-market  rate. 
The  money  market  in  our  sense,  however,  splits  into  so  many  sectors  and 
subsectors  that  no  theoretical  conviction  to  the  effect  that  there  is, 
after  all,  some  meaning  to  the  concept  of  a  single  rate  of  pure  interest 
also  in  this  market,  avails  against  the  fact  that  it  is  impossible  to  indicate 
it  in  a  satisfactory  manner.  As  we  have  seen  above,  we  may  schematize 
those  sectors  and  subsectors  according  to  types  of  borrowers — the  German 
peasant,  for  example,  had  hardly  any  sources  of  balances  to  turn  to  other 
than  saving  banks,  cooperative  banks  (Genossenschaftskassen  of  the 

1  The  late  Alfred  Beit,  the  South  African  financier,  used  to  say  that  he  never  lost  so 
much  on  anything  a§  he  did  pn  consols, 


THE  RATE  OF  INTEREST  617 

Raiffeisen  kind),  and  village  usurers;  to  German  governments  and 
municipalities  any  source  was  open — according  to  types  of  funds — certain 
household  as  well  as  institutional  funds  were  even  by  law  directed  into 
certain  channels;  and  large  classes  of  savers  habitually  turn  to  certain 
kinds  of  investments  and  not  to  others — and  according  to  types  of  title — 
from  the  standpoint  of  the  issuing  corporation,  as  from  the  standpoint  of 
theory,  stocks  and  bonds  are  fundamentally  the  same  thing  and  yet, 
owing  to  the  difference  in  the  rights  they  embody,  they  suit  different 
people  and  different  situations.  Markets  of  a  higher  order  of  specializa- 
tion, in  fact,  develop  around  every  issuing  house  or  financial  group  as 
well  as  around  every  industry. 

In  briefly  surveying  some  of  the  more  important  of  these  special 
markets  and  the  interaction  of  the  rates  that  rule  in  each  of  them,  we 
will  discard  the  market  (which  again  splits  into  many  subsections)  of 
loans  to  households,  except  loans  to  "public  households.'*  As  has  been 
already  mentioned,  however,  an  element  of  lending  to  private  households 
for  purposes  of  consumption  is  included  in  agricultural  credit.  Also, 
we  will  not  specially  deal  with  that  mass  of  credit  that  arises  from  the 
relations  between  manufacturers,  wholesalers,  retailers,  and  households, 
though  not  all  of  it  is  financed  by  banks.  And  we  shall  finally  neglect 
the  problem  of  regional  differences. 

2.  Many  authors,  notably  German  ones,  would  look  for  the  rate  of 
interest  ruling  in  a  country  (landesublicher  Zinsfuss:  what  a  suggestion 
of  traditionality,  ominous  for  our  purpose,  this  turn  of  phrase  conveys!) 
to  the  interest  charged  on  rural  and  urban  mortgage  loans.  There  can, 
however,  be  an  approximation  to  truth  in  this  view  only  in  the  case  of 
first  mortgages  on  urban  realty.  Other  than  first  mortgages  on  urban 
realty  are  in  many  cases  pure  speculation.  As  to  other  than  urban 
realty  we  must  bear  in  mind,  on  the  one  hand,  that  in  many  countries 
habits  of  certain  classes  of  savers,  as  well  as  legislation  regulating  invest- 
ment of  trustees'  funds  and  investment  by  saving  banks,  favor  this  kind 
of  credit,  while,  on  the  other  hand,  mortgages  on  farm  land  imply  a 
special  kind  of  risk,  very  deterrent  to  many  people,  i.e.,  the  risk  of  having 
to  undertake  the  management  of  the  farm  or  to  find  a  manager  for  it. 
These  facts  go  far  toward  setting  up  a  special  market  displaying  a  special 
rate.  In  any  case,  we  cannot  look  for  a  true  representation  of  the  cyclical 
behavior  of  the  rate  to  those  mortgage  rates  that  are  very  sluggish  from 
year  to  year,  except  in  the  case  of  the  Kondratieff. 

Of  course,  though  some  reservations  are  necessary  owing  to  the  organ- 
ization1 of  this  kind  of  credit,  these  rates  do  eventually  follow  the  general 

1  This  organization  perfected  to  the  utmost  in  Germany  during  the  last  20  or  30  years 
of  our  period  and  even  before  that  highly  efficient,  accounts,  for  example,  for  the  fact  that 
the  rise  in  interest  that  begins  in  1895  asserted  itself  but  little  in  this  market. 


618  BUSINESS  CYCLES 

march  of  things.  There  are,  moreover,  two  major  links  with  other  sec- 
tors of  the  money  market.  On  the  one  hand,  only  part  of  the  total 
amount  of  mortgage  credit  consists  in  balances  that  the  lender  intends  to 
tie  up  for  a  long  time.  Other  sources  are  available  and,  at  some  times  and 
in  certain  countries,  particularly  in  the  United  States,  banks  have  gone 
much  further  in  extending  credit  in  those  directions  than  principles  of 
classic  banking  practice — and  common  sense — warrant,  so  that  mortgage 
money  coming  from  nonbanking  sources  often  competes  with  the  ordinary 
sources  of  credit  provided  by  commercial  banks.1  On  the  other  hand, 
mortgages  may  be  "mobilized"  like  every  other  instrument  of  credit, 
and  then  they  invade  the  bond  market  and  establish  connection  with 
the  latter's  rate.  The  most  perfect  instance  of  this  is  the  Pfandbrief  of 
the  German  mortgage  banks  (Hypothekenbanken)  or  of  cooperative  credit 
institutions  of  the  type  of  the  Landschaften.  The  actual  rate  of  interest 
for  this  kind  of  credit  was  not  simply  the  rate  which  the  borrower  con- 
tracted to  pay  but  that  rate  taken  with  reference  to  the  price  at  which 
the  bonds  sold  in  the  market2  and  also  to  the  charge  made  by  the  insti- 
tutions for  their  services.  We  have  various  series,  especially  since  1870,3 
which  very  clearly  bring  out  the  fact  that  at  this  point  mortgage  credit  of 
the  agricultural  and  urban  kind  merges  into  the  general  bond  market. 
In  view  of  what  is  to  follow  later,  we  may  note  that  interest  on  "true" 
mortgages — distinguished  from  mortgages  that  only  secure  other  types 
of  credit — as  reflected  in  the  rate  and  yield  of  Pfandbriefe  (the  writer 
prefers  not  to  translate  the  term)  fell  from  1870-1895  and  then  rose  till 
1914,  although  not  very  much.  The  Preussische  Central-Bodenkredit- 
Aktiengesellschaft,  for  example,  which  offered  its  Pfandbriefe  publicly, 
issued  in  the  year  of  its  foundation  (1870)  a  5  per  cent  type  at  par  (which 
means  5^  per  cent  for  the  borrower).  In  1879  it  financed  on  the  basis  of 
4-Hz  per  cent  at  par,  in  1884  at  4  per  cent,  in  1889  at  nearly  3j-£  per  cent. 
From  1890  to  1894  it  had  to  return  to  the  4  per  cent  type.  It  reached  the 
minimum  in  1895,  after  which  a  slow  and  moderate  rise  set  in  (see  Horn- 
burger,  table  on  page  97). 

3.  The  bond  market  is  another  semiindependent  piece  of  the  mecha- 
nism which  we  call  the  money  market.     The  fences  that,  worn  away  in 

1  The  reader  will,  however,  recall  that  banks,  when  they  lend  for  agricultural  purposes 
other  than  current  business  or  for  urban  building,  do  not  always  commit  that  mortal  sin 
against  the  classic  rules  of  banking  which  has  so  much  to  do  with  difficulties  in  depression. 
Besides,  it  has  been  mentioned  already  that  mortgages  may  be  no  more  than  an  additional 
security  for  an  unexceptionable  credit. 

2  This  calculation  presents  various  difficulties  into  which  we  cannot  enter  here.     In 
particular,  it  would,  in  case  the  bonds  are  above  par,  be  correct  only  if  the  whole  premium 
went  to  the  borrower,  which  in  practice  was  not  generally  the  case. 

3  See  Hecht,  Organisation  des  Bodenkredits  in  Deutschland,  statistics  of  mortgage 
banking  in  Part  II,  vol.  1 ;  also  Homburger,  cited  before,  p.  68,  et  seq.;  for  older  data,  see 
Von  Cyriaci-Wantrup,  Agrarkrisen  und  Stock ungsspannen,  p.  87. 


THE  RATE  OF  INTEREST  619 

many  places  but  still  recognizable,  separate  it  from  other  sections  consist 
not  only  of  costs  and  risks  but  also  of  habits  of  investors  and  positions  of 
borrowers.  Their  reality  is  proved  by  the  fact  that,  in  the  short  run, 
bond  yields  sometimes  differ  considerably  not  only  from  "short"  but 
also  from  other  "long"  rates.  But  the  bond  market  is  also  very  liable 
to  split  into  subsections — so  much  so  that  different  issues  of  the  same 
borrower  occasionally  display  an  individualism  (see,  for  example,  the 
different  behavior  during  the  20  years  before  the  war  of  the  German 
government  3  and  3%  per  cents)  that  is  not  always  easy  to  explain. 

One  important  subsection  the  individuality  of  which  is  accentuated 
by  legal  privileges  of  various  kinds  and  by  a  strong,  if  sometimes  uncrit- 
ical, preference  on  the  part  of  savers  in  some  countries — particularly  in 
France — is  constituted  by  government  bonds.  In  England  and  Ger- 
many, however,  this  market  received  a  shock  by  the  policy  of  conversions, 
which  in  part  at  least  sprang  from  an  erroneous  diagnosis  of  the  nature  of 
the  general  fall  in  rates  to  1897  and  possibly  also  from  an  overestimation 
of  immediate  gain  as  against  future  disadvantage.  In  the  United  States, 
confidence  in  the  currency  established  in  1879  or,  practically,  in  1878,  was 
repeatedly  shaken  by  soft-money  agitation  and  free-silver  campaigns, 
one  of  which  is  perhaps  responsible  for  the  small  peak  in  yields  that 
occurred  in  1896,  when  the  yield  of  the  United  States  4  per  cents  increased 
by  J4  per  cent.  But  the  policy  entered  upon  in  the  late  eighties  of  buying 
up,  regardless  of  cost,  influenced  quotations  much  more  strongly  in  the 
opposite  direction.  The  fact  that  the  yield  of  United  States  4  per  cents 
fell  till  1901,  unlike  that  of  German  governments  which  touched  the  low 
point  in  1894  and  that  of  English  consols,  which  did  so  in  1897,  *  may 
possibly  be  explained  by  this,  although  other  bonds  did  the  same  thing.2 

1  The  high  point  in  the  London  and  Cambridge  Economic  Service  Index  of  the  prices 
of  fixed  interest  securities,  which,  however,  gives  rise  to  various  doubts,  comes  in  1896; 
the  French  rente  reached  its  maximum  in  1897. 

2  Mr.  Macaulay's  Index  of  American  railroad  bond  yields  (the  "  yield  of  best  bonds  each 
January")  touches,  however,  its  low  point  in  1899  (Journal  of  the  American  Statistical 
Association  for  March  1926.     The  reader  should  refer  to  this  paper  for  questions  of  tech- 
nique and  some  points  of,  comparatively  speaking,  detail).     Moody  *s  monthly  series  of 
corporation  bond  prices,  1866-1914,  described  in  Moody's  Investors'  Service,  November 
1924,  reaches  a  high,  and  the  Standard  Statistics  Index  of  Yields  upon  60  High  Grade  Bonds 
a  low  point,  as  late  as  1902.     The  Treasury's  policy  of  debt  redemption  beyond  sinking 
fund  requirements,  so  vigorously  pursued,  may  in  part  account  also  for  this,  because  a 
considerable  fraction  of  the  funds  set  free  thereby  would  understandably  seek  investments 
similar  in  character.     The  movement,  contrary  to  expectation,  lasts,  however,  exactly  as 
long  as  the  sharp  increase  of  investments  by  National  Banks.     Besides,  debtor  concerns 
certainly  became  better  as  a  rule  after  the  crisis  of  1898  was  over.     There  is  no  need, 
therefore,  for  concluding  that,  as  the  yield  of  governments  behaved  substantially  as  other 
bonds,  we  are  faced  with  a  true  movement  of  "the  long  rate"  asserting  its  independent 
existence  and  indicative  of  deep-seated  causes  in  the  sphere  of  "real  capital." 


620  BUSINESS  CYCLES 

All  this  need  not  detain  us.  The  general  contour  stands  out  clearly. 
As  to  other  than  government  bonds,  the  reader  should,  however,  bear  in 
mind  that  in  this  country  the  market  of  prewar  days  differed  fundamen- 
tally from  that  of  the  twenties  of  this  century  in  material,  in  public,  and 
in  financial  structure.  Railroad  bonds  were  the  big  item.  They  and, 
still  more,  the  bonds  that  the  merger  movement  produced  were  primarily 
taken  in  large  lots  by  institutional  and  wealthy  private  investors.  The 
general  public  had  not  much  to  do  with  them,  and  Europe  gradually  sold 
out.  In  England,  also,  railway  debentures  were  the  most  prominent 
"business"  element  of  the  bond  market,  which  dealt  mainly  in  local, 
municipal  and  quasi-municipal,  and  foreign  loans.  In  Germany  a  market 
in  industrial  as  distinguished  from  railway  bonds,  important  enough  to 
figure  and  to  display  regularities  of  its  own,  developed  in  the  course  of  the 
eighties.  Many  factors,  international  influences,  and  changing  quality 
or  security  or  changing  financial  standing  of  debtors,  among  others, 
would  have  to  be  taken  into  account  in  order  to  explain  the  behavior  of 
rates  in  this  market  and  the  changing  relations  between  the  rates  ruling 
in  its  various  sectors.  For  us,  however,  it  is  sufficient  to  complement  our 
statements  at  the  end  of  the  preceding  section  with  a  few  remarks  on  the 
mechanism  which  links  bond  yield  to  short  money  rates. 

In  prewar  times,  and  particularly  in  London,  speculators  of  a  certain 
type — they  were  really  a  kind  of  arbitrageur — reacted  promptly  to  very 
small  variations  in  the  relation  between  bond  yields  and  short  rates,  and 
at  once  borrowed  short  in  order  to  buy  bonds  when  the  margin  covered 
cost  and  risk.  In  this  country  the  negative  short-time  association 
between  the  loans  and  the  investments  of  banks  outside  New  York  City, 
to  be  mentioned  again  in  the  next  chapter,  is  indicative  of  another  link. 
And  the  financing  of  subscriptions  to  bond  issues  by  bank  loans  and  the 
refunding  of  bank  loans  from  the  proceeds  of  bond  issues  provides  a 
third.  We  readily  see  not  only  that  this  mechanism  is  by  nature  inca- 
pable of  equalizing  bond  yield  and  short  rates,  owing  to  the  particular 
kind  of  commitment  incident  to  the  legal  construction  of  the  claim 
embodied  in  a  bond,  but  also  that  for  reasons  of  financial  technique  it 
will  be  incapable  of  enforcing  instantaneous,  still  less  proportional, 
covariation.  Even  cases  of  variations  in  the  opposite  direction  are  by  no 
means  rare.  For  instances  we  need  not  look  only  to  situations  of  deep 
depression,  in  which,  when  the  panic  is  over,  short  rates  may  drop  almost 
to  vanishing  point,  while  bond  rates  and  bond  yields  stay  up.  It  can 
happen  in  the  most  normal  course  of  things  that,  for  example,  when  a 
large  issue  has  been  negotiated  and  the  borrower  next  lends  the  proceeds 
in  the  open  market,  bond  yields  rise  and  short  rates  fall  for  the  time  being. 
International  relations,  though  we  do  not  now  take  them  into  account, 
may  be  more  effective  in  the  one  market  than  they  are  in  the  other.  A 


THE  RATE  OF  INTEREST  621 

sufficiently  widespread  tendency  among  industrial  borrowers  to  make 
themselves  independent  of  their  banks  may  occasionally  raise  the  one  and 
depress  the  other  rate.  It  also  should  be  clear,  however,  in  what  sense 
we  may  say  that  all  this  is  "merely  technical."  In  spite  of  it,  the  bond 
market  is  fundamentally  but  the  most  sluggish  part  of  the  money  market, 
and  in  its  rates  reflects,  though  sometimes  with  a  lag  and  always  at  a 
higher  level  and  with  smaller  amplitudes,  the  fluctuations  in  short  rates, 
which  not  only  express  business  situations  more  faithfully  because  they 
are  free  to  react  instantaneously,  but  also  are  the  main  influence  in 
shaping  bond  yields.  The  behavior  of  the  two  rates  relatively  to  each 
other,  in  fact,  bears  out — although,  by  itself,  it  does  not  prove — the 
consequences  that  follow  from  this :  it  has  for  this  country  been  shown  by 
Professor  W.  M.  Persons  that  the  graph  of  bond  yield,  plus  one-half  of 
one  per  cent,  looks  strikingly  like  a  "trend  line"  drawn  through  the  graph 
of  short  rates.1  Deviations  from  this  evidently  basic  relation  are  dis- 
tinctly secondary  and  can  easily  be  accounted  for.2 

The  problem  of  stock  prices  will  be  dealt  with  in  the  next  chapter. 
Here  it  is  sufficient  to  repeat  that,  according  to  the  view  taken  in  this 
book  of  the  shareholders'  position  and  of  the  nature  of  their  dividends, 
the  stock  market  is  simply  another  section  of  the  general  market  of  bal- 
ances. Since  profit  is  for  us  the  pillar  of  interest,  dividends  per  cent  of 
capital  actually  paid  in,  which  bear  a  particularly  close  relation  to  profits, 
are  really  the  most  fundamental  of  all  the  rates  of  the  money  market. 
Broadly  and  on  the  average,  they  move  in  the  same  direction  as  bond 
yields  (or  rates  on  new  bonds)  and  short  rates.  But  the  dominant 
influence  of  those  risks  and  chances  that  are  peculiar  to  common  and, 
to  a  lesser  degree,  to  all  stock,  impairs  the  working  of  the  mechanism  that 

1  Compare  his  study  in  the  Review  of  Economic  Statistics  for  April  1927,  chart  on  pp.  94 
and  95,  see  Chap.  V,  Sec.  B,  where  this  was  given  as  an  instance  for  a  "reference  trend." 
Crossing  points  roughly  coincide  with  the  inflection  points  of  a  smoothed  curve  of  short 
rates  and  are — again,  roughly — indicative  of  the  Kitchin  cycles.     Otherwise  they  are  not 
easy  to  interpret.     There  is,   in  particular,   "not  always  a  one-to-one  correspondence 
between  crossing  points,  on  the  one  hand,  and  peaks  and  troughs  (of  the  curves  for  four 
series  of  security  prices)  on  the  other  hand";  nor  is  there  any  "pronounced  tendency 
revealed  for  crossing  points,  on  the  one  hand,  and  peaks  and  troughs,  on  the  other  hand, 
to  occur  simultaneously"  (op.  cit.,  p.  101). 

2  See  Dr.  C.  E.  Thomas's  paper,  The  Effect  of  the  Depression  upon  Bond  Yields; 
Journal  of  the  American  Statistical  Association,  September  1988.     His  result  (though  arrived 
at  mainly  from  postwar  material)  that  in  an  approximately  normal  situation  a  decrease 
of  1  per  cent  in  the  short  rate  would  produce  a  decline  of  approximately  0.24  per  cent  in 
bond  yields  (p.  267)  is  as  suggestive  as  is  his  emphasis  on  the  influence  of  trade  volume,  any 
decline  of  which  tends  to  counteract  the  influence  of  a  fall  in  short  rates  and  to  reduce 
it  practically  to  zero  precisely  in  "deep"  depression  (p.  271).     It  is  important  to  bear  in 
mind  that  this  does  not  contradict  any  of  the  statements  in  our  text.     Volume  of  Trade 
influences  revenue,  revenue  influences  the  safety  of  the  bondholder's  investment  (op.  cit., 
p.  262). 


BUSINESS  CYCLES 

links  this  section  of  the  market  to  the  other  two.  Neither  investors  nor 
"borrowers"  can  choose  between  purchasing  and  issuing  bonds  or  stocks 
according  to  actuarial  rationality  alone.  And  there  is  no  simple  and 
reliable  relation  between  money  rates  and  stock  prices  such  as  there  is — 
barring  risk  of  default — between  money  rates  and  bond  prices.  The 
primarily  negative  association  between  stock  and  (highest  grade)  bond 
prices,  however,  not  only  indicates  that  that  mechanism  is  after  all  not 
absent,  but  also  reveals  the  way  in  which  the  counteracting  influences  of 
variations  in  returns  (dividends)  and  variations  in  pure  interest  work  out 
in  the  course  of  cyclical  situations.1 

4.  In  one  way  or  another,  all  the  types  of  transactions  in  balances 
that  we  have  so  far  discussed  thus  lead  toward  the  Open  Market.  This 
is  contrary  to  expectation,  inasmuch  as  our  schema  has  on  its  monetary 
side  been  made  to  center  in  the  relation  between  member  banks  and  their 
customers.  We  would,  in  fact,  select  for  guide  in  our  analysis  the  interest 
charged  to  customers  in  current  account  (Kontokorrent-Kredit,  cus- 
tomer's line  of  credit),  the  true  bank  loan  rate,  if  we  could.  From  this, 
however,  we  are  not  only  estopped  by  the  impossibility  of  compiling  a 
reliable  series  for  prewar  times,  but  also  by  those  difficulties  inherent  in 
this  material  of  which  mention  has  been  made  above.  This  market  is  so 
far  from  perfect,  charges  differ  so  much  not  only  locally  but  also  as 
between  banks,2  customers  and  transactions,  and  are  so  sluggish,  that 
tendencies  can  be  read  off  much  better  from  the  series  of  the  New  York 
commercial  paper  rate3  or,  for  England  and  Germany,  from  other  open- 
market  rates  which  in  those  countries  more  or  less  correspond  to  that 
American  series.  This  choice  is  not  satisfactory.  The  open  market 
and  its  immediate  vicinity,  indeed,  enjoy  greater  (almost  absolute) 
freedom  from  those  influences  that  make  for  stickiness  and  are  almost 
perfectly  competitive.  But  they,  more  than  other  sectors,  are  exposed 
to  the  regulative  activity  of  central  banks — degree  and  method  of  which 
greatly  varied  during  the  long  period  under  discussion — and  to  the  dis- 
ruptive influence  of  the  factor  which,  especially  in  England,  it  was  the 
principal  aim  of  that  activity  to  control.  This  factor  is  the  mass  of 

1  Cf.  W.  M.  Persons,  Money  Rates  and  Security  Prices,  Review  of  Economic  Statistics, 
January  1926. 

2  In  some  countries,  particularly  in  Germany,  this  situation  produced  the  result  that 
is  so  often  the  outcome  of  market  imperfections:  banks  entered  into  agreements  with  each 
other  (Konditionenkartelle)  with  a  view  to  normalize  charges.     But  this  affects  the  last 
prewar  decade  only. 

8  To  the  published  figure  between  ^  and  1  per  cent  should  be  added,  in  order  to  arrive 
at  the  rate  actually  paid  by  borrowers.  The  remuneration  for  the  series  of  intermediate 
dealers  is,  in  general,  not  comprised  in  published  open-market  rates.  The  relation  between 
the  rate  on  60  to  90-day  paper  and  4  to  6  month  paper  is  as  we  should  expect,  and  does  not 
call  for  comment. 


THE  RATE  OF  INTEREST  623 

temporarily  unemployed  balances  and  credit  facilities,  which  in  that 
market  hunted  "for  any  wretched  rag  of  interest." 

It  is  here  that  what  we  have  called  Temporary  Investment  puts  in 
appearance.  Funds  which  large  concerns  have  accumulated  for  later  acts 
of  real  investment  or  simply  as  a  cash  reserve  or  which  they  have  on  their 
hands  when  they  are  underspending  in  depression,  or  surplus  funds  of 
public  or  semipublic  bodies,  such  as  the  Prussian  state  railroads  and  the 
Prussian  government  or  India  House,  flowed,  with  increasing  rationali- 
zation in  the  management  of  funds,  increasingly  into  the  open  market  for 
normally  riskless  and  ideally  short  investment.  In  these  cases,  banks 
were  frequently  employed  as  agents.  But  they  did  a  similar  thing  on 
their  own  account  if  their  own  credit  facilities  were  underutilized  in  their 
own  business  and  also  in  order  to  collect  their  secondary  reserve  (see 
Chap.  XIII).  They  even,  especially  in  England,  shifted  to  the  open 
market  part  of  what  otherwise  would  have  been  their  normal  business 
with  their  own  customers :  bill  brokers  (discount  houses)  looked  after  the 
acceptors  of  that  part  of  their  portfolio  and  relieved  them  of  much  trouble; 
this  business  was  impersonal  and  restriction  was  normally  possible  at  any 
moment  without  loss  of  either  money  or  prestige  and  without  remon- 
strances from  irate  customers;  while  firms  such  as  the  merchants  of 
London  and  the  great  provincial  centers  were  glad  to  offer  their  remit- 
tances in  Lombard  Street,  seeing  that  as  a  rule  they  got  lower  rates  from 
their  broker  than  they  would  have  from  the  very  bank  that  financed  that 
broker.  This  means,  of  course,  that  the  amount  of  actual  and  potential 
balances  offered  in  that  market,  while  very  variable,  is  extremely  inelastic. 
No  fall  in  rates  discourages  it;  save  in  exceptional  circumstances,  such  as 
obtain  during  speculative  manias,  it  hardly  increases  in  response  to  a  rise 
except  by  way  of  immigration  of  foreign  balances.1  This,  together  with 
the  perfect  competition  which  rules  in  the  open  market,  accounts  for  the 
frequency  and  the  amplitude  of  variations  in  its  rates,  as  well  as  for  the 
close  covariation  of  all  of  them. 

All  this  undoubtedly  deprives  these  variations  of  much  of  the  signifi- 
cance they  would  otherwise  have.  It  should  be  remembered,  however, 
that  this  significance  is  not  further  reduced  by  the  short-time  character  of 
the  operations  which  those  funds  serve.  For  we  know  that  this  character 
is  largely  delusive :  at  least  through  financing  the  carrying  of  stock,  even 
call  loans  are  by  no  means  debarred  from  financing  such  things  as  railroad 
building,  while  other  forms  of  open-market  credit  much  more  directly 
serve  the  purpose  of  long-time  real  investment.  Finally,  we  have  to  add 
to  our  list  of  reasons  why  rates  of  interest  should  lag  behind  other  ele- 

1  More  correctly  and  also  more  generally :  by  way  of  immigration  into  the  banking  circle 
under  consideration;  for  large  countries  may  so  divide  up  into  sectors  that  a  rising  rate 
may  be  effective  in  drawing  floating  funds  from  one  to  another. 


624  BUSINESS  CYCLES 

ments  in  the  cyclical  process  one  more  item  that  is  peculiar  to  open-market 
rates,  which  are  so  remarkably  free  from  the  influence  of  others.  As  far 
as  the  balances  floating  in  the  open  market  have  their  source  in  under- 
spending, they  obviously  cannot  act  on  rates  until  restriction  of  indus- 
trial and  commercial  operations  has  occurred  to  a  significant  extent,  that 
is  to  say,  when  liquidation  has  already  turned  into  depression.  Then 
the  effect  of  decreasing  borrowing  is  intensified  by  the  increase  in  funds 
offered.  The  causal  importance  of  this  for  the  puls"e  of  business  is, 
however,  small,  as  we  shall  presently  see.  And  the  location  in  time  of  the 
upper  turning  point  in  the  money  curve  of  the  Harvard  Index  Chart  is  not 
as  a  rule  explained  by  it.  But  the  lag  of  the  lower  turning  point  in  the 
interest  series  may  be,  for  until  the  slack  due  to  underspending  is  taken 
up,  open-market  rates  cannot  recover  from  their  trough. 

Call  (broker's  loans'  renewal)  rate  and  time  rate  do  not  differ  more 
than  we  should  expect  from  the  smoothing  effect  the  time  arrangement 
must  have.  Commercial  paper  rate  varies  much  as  time  rate,  though  at  a 
somewhat  higher  level.  No  systematic  change  in  the  relation  between 
the  two  seems  to  have  occurred  in  this  country  or  in  Germany  during  the 
last  30  years  before  the  war  and  not  much  in  England,  but  there  is  in  it 
a  characteristic  cyclical  variation.  We  will,  however,  confine  ourselves 
to  a  few  comments  on  some  of  the  instruments  of  credit  that  are  mainly 
financed  by  open-market  balances. 

Treasury  bills — we  use  this  term  for  all  perfectly  negotiable  short 
government  obligations  and  for  all  countries — are,  from  the  standpoint  of 
the  buyer,  merely  the  senior  members  of  the  family  of  bankers'  acceptances 
and  move  with  the  rest  at  a  premium,  which  understandably  widens  in 
prosperities  and  shrinks  in  recessions.  They  may,  for  our  purposes  at 
least,  be  said  to  mean  much  the  same  thing  in  our  three  countries.1  But 
bankers'  acceptances,  and  therefore  their  rates,  do  not.  They  are  first 
the  typical  instrument  for  the  financing  of  current  commodity  trans- 
actions in  international  trade,  and  as  such  they  have,  up  to  the  end  of  the 
seventies,  primarily  belonged  to  the  English  domain.  It  was  not  until 
the  eighties  that  Germany  increasingly  took  to  financing  her  foreign 
trade  herself.  Something  of  that  close  relation  to  individual  commodity 
transactions,  at  first  so  characteristic  for  this  instrument  of  credit,  was 
lost,  even  in  England,  in  the  case  of  the  finance  bills  that  used  to  serve, 
and  still  serve,  in  American  export  trade.  But  English  practice  sub- 
stantially confined  itself  to  this  particular  purpose  and  did  not  knowingly 
use  finance  bills  for  any  other  to  a  significant  extent. 

In  America  bankers'  acceptances  of  this  type,  and  a  market  for  them, 
are  a  war  and  postwar  growth.  Before  the  war,  accepting  was  even  ultra 

1  The  Austrian  Salinenscheine  afford  the  first  European  instance  known  to  the  writer 
of  open-market  operations  undertaken  by  a  government  for  purposes  of  regulation. 


THE  RATE  OF  INTEREST  625 

vires  for  National  Banks.  But  in  Germany  bankers'  acceptances  were 
also  used  to  a  very  large  extent  (much  diminished  after  the  war)  for 
purposes  of  domestic  trade  and  not  only  for  that  but  even  as  a  means  of 
raising  funds  for  real  investment  and  of  financing  plant  and  equipment. 
Thus,  the  bankers'  acceptance,  which  costs  the  bank  nothing  but  its 
signature,  became  so  popular  as  to  acquire  a  central  position  in  the  market. 
A  technique  grew  up  for  trading  in  these  acceptances  which  made  them 
ideally  salable  at  a  single  rate  by  standardizing  the  "commodity"  through 
certain  requirements  as  to  the  standing  of  the  accepting  bank  and  as  to 
form  (round  figures;  observe  the  characteristically  different  attitude  of 
English  banks  in  this  point).  They  were  even  directly  used  as  means  of 
payment.  There  is  no  perfect  parallel  to  this  in  England  or  America, 
but  the  writer  has  come  to  the  conclusion,  in  spite  of  considerable  doubts, 
that  this  rate  (Privat-Diskont)  is  the  nearest  equivalent  to  the  New  York 
commercial  paper  and  the  Lombard  Street  rates.  The  decreasing  impor- 
tance of  the  commercial  bill  must,  however,  be  borne  in  mind:  with 
English  banks,  for  example,  bills  were  26  per  cent  of  assets  in  1880  and 
only  1£  per  cent  in  1912. 

5.  Finally,  it  must  be  borne  in  mind  that  although  in  our  schema 
bankers'  banks  and  their  rates  stand  "at  one  remove"  from  the  money 
market,  this  was  not  so  historically.  Not  until  the  Federal  Reserve  Act 
did  any  bankers'  bank  confine  itself  to  bankers'  bank  business  only. 
On  the  contrary,  they  all  did  member  banks'  business  also — at  times, 
primarily — and  their  central  positions  and  bankers'  bank  policies  grad- 
ually evolved  from  it.  In  the  case  of  the  New  York  banks,  to  which, 
nevertheless,  we  feel  justified  in  attributing  central  bank  functions,  this 
precludes  any  search  for  a  bank  rate  in  the  sense  of  our  analysis.  The 
Bank  of  England's  rate  was,  in  the  times  of  the  bill  and  the  bank  note, 
simply  the  rate  at  which  it  was  prepared  to  do  business  precisely  with 
nonbanking  customers.  This  rate  was  highly  inflexible  (between  4  and 
5  per  cent)  for  over  a  century,  until  Peel's  Act  (except  in  the  latter  part  of 
1838).  Then  what  was  called  the  New  System  of  Discounting  was 
embarked  upon.  It  consisted  in  competing  for  paper  in  the  market  and 
resulted  for  a  time  in  a  dominating  position  of  the  Bank  in  this  business. 
Rate  policy  was  currently  adapted  to  that  purpose,  and  although  it  would 
not  be  correct  to  say  that  the  Bank  merely  followed  the  market — a  buyer 
who  acquires  50  per  cent  of  any  commodity  can  never  be  said  merely  to 
"follow,"  (see  the  next  chapter) — its  rate  was  certainly  just  one  of  the 
market  rates. 

Up  to  1858  the  Bank  at  least  rediscounted  for  bill  brokers  and  dis- 
count houses,  thus  doing  typical  bankers'  bank  business,  which  practi- 
cally amounted  to  rediscounting  for  banks  in  the  usual  sense  of  the  word1 
1  Bill  brokers  and  discount  houses,  it  should  be  recalled,  are  for  us  a  special  type  of 
bankers  or  "satellites"  of  bankers.  What  is  meant  above  is  that  the  London  banks  which 


626  BUSINESS  CYCLES 

and  in  effect  overcame  the  It  Is  Not  Done  which  obstructed  the  march 
of  the  latter  to  the  Bank.  But  in  that  year  rediscounting  for  the  bill 
market  was  discontinued,  and  the  Bank  thenceforth  entirely  relied  on 
its  ability  to  attract  the  desired  amount  of  paper  competitively1  until  the 
early  seventies,  when  this  practice  ceased  to  give  satisfaction  to  the 
public.  Until  then  bank  rate,  though  as  a  rule  on  a  higher  level,  would 
serve  our  purpose  almost  as  well  as  market  rate.  But  after  some  years  of 
hesitation,  the  Bank  then  gave  the  strongest  proof  of  its  determination 
to  cling  to  its  member  bank  business.  In  1878  it  declared  that  it  would 
no  longer  adhere  to  its  official  rate  when  discounting  for  its  own  cus- 
tomers but  charge  them  about  the  market  rate.  At  the  same  time,  it 
made  a  move  to  retrace  its  steps  in  the  matter  of  rediscounts  by  inti- 
mating its  willingness  to  make  special  advances  to  bill  brokers  (discount 
houses),  although  rediscounting  for  them — at  the  official  rate — was  not 
resumed  until  1890.2  These  two  measures  are  highly  significant  precisely 
because  they  were  taken  together.  They  mark  a  new  stage  in  the  evolu- 
tion of  the  Bank's  policy:  while  it  asserted  its  member  interest,  it  simul- 
taneously divorced  this  from  its  central  bank  function;  bank  rate  ceased 
to  be  the  hybrid  that  it  had  been  until  then,  and  began  to  assume  the 
place  assigned  to  it  in  our  schema. 

Similarly,  the  rate  of  the  Prussian  and  later,  the  German  Imperial 
Bank  was  primarily  a  discount  and  only  secondarily  a  rediscount  rate. 
Although  the  Bank  was  always  under  pressure  from  popular  demands  for 
a  cheap  and  especially  steady  rate  and  at  times  had  to  yield  to  them,  we 
may  repeat  for  Germany,  what  has  been  said  in  the  case  of  the  Bank  of 
England,  that,  for  the  nineteenth  century  at  least,  we  could  do  almost 
as  well  with  bank  rate  as  with  any  market  rate. 

C.  Discussion  of  the  Time  Shape  of  Interest  Rate. — This  discussion 
will  primarily  be  based  on  the  series  presented  in  the  pulse  charts,  which 
also  show  how  interest  behaves  relative  to  other  elements  of  the  cyclical 
process  (see  also  Chart  XX) .  Call  rate  (Charts  XXXVI  and  XXXVII), 
in  its  relation  to  stock  exchange  figures,  will  be  briefly  discussed  in  the 

financed  brokers  were  enabled  to  call  their  loans  at  will,  since  the  latter  could  go  to  the 
Bank,  and  that  this  put  them  in  much  the  same  position  as  if  they  had  discounted  and  then 
rediscounted  themselves. 

1  Under  conditions  of  imperfect  competition  that  is,  especially  in  England,  entirely 
compatible  with  being  dearer  than  one's  competitors.     In  its  provincial  branches,  however, 
the  Bank  discounted  at  the  prevailing  local  rate  and  sometimes  below  it. 

2  Mr.  King  in  History  of  the  London  Discount  Market,  p.  304,  stresses  the  later  rather 
than  the  earlier  date  and  thus  misses  the  essential  complementarity  of  those  two  measures. 
He  even  (p.  297  n.)  accuses  Palgrave  of  inaccuracy  in  holding  that  the  policy  of  1858  was 
reversed  in  1878.     But  Palgrave  does  not  seem  to  assert  much  more  than  Mr.  King  admits. 
The  difference  seems  to  arise  from  an  inclination  of  Mr.  King  to  underline  the  achievements 
of  William  Lidderdale  very  strongly.     His  own  interpretation  of  the  resumption  of  redis- 
t, however,  impaired  thereby. 


THE  RATE  OF  INTEKEST 

next  chapter.  From  the  outset  it  must  be  borne  in  mind  that  the  rates 
chosen,  although  comparatively  "pure,"  still  display  the  influences  of 
legislative  changes,  currency  and  credit  policies,  gold  production,  inter- 
national relations,  and  in  particular  of  panics — all  of  which  unavoid- 
ably interfere  with  expected  behavior  and  with  such  barometric  value  as 
the  rate  of  interest  may  be  held  to  have.  But  a  barometer  that  can 
be  spoiled  is  still  a  barometer,  and  in  large  intervals  within  our  period 
some  of  these  influences — though  not  that  of  panics,  which  acted  on 
interest  much  more  strongly  than  they  did  after  the  World  War — were 
not  very  important. 

1.  Since  we  do  not  use  indices  of  rates,  interest  series  are  natural.  Of 
course,  they  are  systematic — in  fact,  they  may  justly  be  called  the  most 
systematic  of  all.  There  cannot  be  any  doubt  that  what  they  describe 
is  a  primary  element  of  the  cyclical  process  of  evolution,  though  not  all 
economists  will  agree  to  the  statement  that  follows  from  our  analysis,  viz., 
that  interest  is  fundamentally  consequential,  and  is  causal  only  in  a 
secondary  sense.  Deferring  additional  comment  on  the  latter  point,  we 
note  that,  similarly,  everybody  will  accept  the  proposition  that  interest 
is  the  most  cyclical  of  all  the  elements  of  the  system;  but  hardly  anybody 
the  further  proposition  that  it  makes  a  "clean"  cyclical  series.  Improve- 
ment of  the  organization  of  credit,  decreasing  premia  for  risk,  and  other 
such  circumstances  may  account  for  a  falling  descriptive  trend  in  interest 
series,  but  our  theory  denies  that  there  is  any  result  trend.  While  in  the 
pure  model  interest  would  in  each  cycle  start  from  and  return  to  zero, 
there  is  no  generally  valid  reason  why  it  should  not  in  actual  fact  come  as 
near  to  it  in  any  neighborhood  of  equilibrium — account  being  taken  of 
the  phases  of  underlying  cycles — as  it  does  in  any  other.1 

Data  good  enough  to  test  that  proposition  under  comparable  condi- 
tions over  a  sufficiently  extended  period  do  not  and  cannot  exist.  But 
such  information  as  we  have  about  centers  of  capitalist  evolution  anterior 
to  our  period  is  corroboratory  rather  than  otherwise.  In  Amsterdam, 
market  rate  in  the  eighteenth  century — even  in  the  seventeenth — was 
sometimes  as  low  as  2  per  cent.2  However  hazardous  the  comparison — 

1  There  might  be  a  special  reason,  however,  incident  to  the  factor  of  consumers'  borrow- 
ing: the  influence  of  bourgeois  mentality  which  had  created  a  parsimonious  state,  being 
displaced  by  the  influence  of  another  mentality,  making  for  an  extravagant  state,  may 
produce  a  noncyclical  rise  in  interest  in  the  future  or,  as  an  alternative,  currency  disorders. 
Such  conditions  actually  did  prevail  before  the  reign  of  the  bourgeoisie  and  so  did  the  conse- 
quence envisaged,  while  the  opposite  consequence  actually  ensued  as  those  conditions 
passed.     There  may  also  be  spurious  reasons:  as  pointed  out  in  the  text,  modern  technique 
does  not  allow  interest  to  soar  to  panic  peaks. 

2  Even  in  the  England  of  Walpole  and  Pelham,  8  per  cent  consols — the  price  of  which 
we  could  not,  however,  for  reasons  stated  in  Sec.  B,  take  as  a  reliable  guide — occasionally 
sold  at  prices  higher  than  any  they  ever  reached  afterward  (maximum,  107),  see  W.  St. 
Jevons's  chart  in  his  volume  on  Currency  and  Finance. 


BUSINESS  CYCLES 

in  actual  fact,  minima  cannot  always  be  taken  to  indicate  neighborhoods 
of  equilibrium,  see  infra — it  is  not  without  significance  that  in  1897 
London  market  rate  was  about  as  high  (average  for  that  year,  1  pound, 
18  shillings,  6  pence).  And  again  the  question  seems  in  order  as  to  how 
much  the  reader  thinks  would  have  been  left,  in  either  case,  of  those  2  per 
cent  if  there  had  been  no  consumers'  borrowing.  Moreover,  inspecting 
the  graph  of  our  series  of  New  York  commercial  paper  rates  from  1875 
to  1914,  we  find  that  it  fluctuates  about  what  in  a  first  approximation 
seems  to  be  a  perfectly  even  level.  In  fact,  it  looks  as  if  a  trend  had  been 
taken  out  of  it  which  may,  however  short  the  series,  be  adduced  in  favor 
of  our  thesis,  because  it  is  obviously  not  likely  that  if  a  systematic  tend- 
ency to  fall  were  present  in  interest  rates,  this  tendency  would  not  have 
shown  at  all  within  40  years  of  both  vigorous  and  relatively  undis- 
turbed capitalist  evolution.1  It  should  be  observed  in  passing  that  these 
facts,  by  refusing  support  to  the  so-called  Law  of  the  Declining  Rate  of 
Interest  or,  as  the  classics  said,  Profits,  indirectly  also  cast  doubt  on  the 
theories  of  which  that  law  is  a  consequence.  Nor  should  we  be  surprised 
at  this.  There  is  nothing  in  theory  or  fact  to  justify  belief  in  any  tend- 
ency in  the  recurrent  waves  of  profit  (in  our  sense)  either  to  increase  or  to 
decrease  systematically.  It  follows  that  interest  fundamentally  deriving 
from  profit  should  also  not  display  such  a  tendency. 

2.  The  reader  will  readily  see  that  expectation  as  to  the  cyclical 
behavior  of  interest,  if  we  adapt  it  to  the  higher  approximations  of  our 
model,  including  the  coexistence  of  three  cyclical  movements,  will  be  for  a 
lagged  rise  in  prosperity  and  a  similarly  lagged  fall  in  recession.2  Since 
the  neighborhood  value  of  interest  rate  is  not  actually  zero,  further  fall 
will,  in  general,  occur  in  depression  as  a  consequence  of  abnormal  restric- 
tion of  volume  of  business.  But  as  in  all  other  cases,  it  must  be  remem- 
bered that  the  processes  of  depression  are  erratic,  even  apart  from  the 
effects  of  panics  on  the  rate  of  interest  which  are  peculiar  to  it.  Uncer- 
tainty extends,  in  consequence,  to  its  behavior  in  revival,  which  should 
bring  it  back  to  neighborhood  value.  Again  it  must  be  emphasized  that 
these  expectations  will  work  out  somewhat  differently  in  cycles  of  different 
span. 

1  Mr.  Carl  Snyder,  Interest  Rate  and  Business  Cycle,  American  Economic  Review, 
December  1925,  p.  697,  notes  that  "it  is  striking  how  often  this  peak  [of  cyclical  expansion 
as  measured  by  the  Clearings  Index  of  the  Federal  Reserve  Bank  of  New  York]  has  come 
close  to  the  time  at  which  the  interest  rate  has  risen  to  just  above  the  average  of  the  half 
century,  i.e.,  when  the  interest  rate  crosses  the  average  line"  (which  is  at  4.98  per  cent,  1875 
to  1925).     This,  again,  could  not  be  if  there  had  been  a  significant  trend. 

2  The  writer  has  sometimes  heard  and  read  the  statement  that  the  fluctuations  of  credit 
lag  behind  fluctuations  in  business  activity.     This  is  not  the  same  thing — and  not  quite 
correct — although  it  is  obvious  that  the  statement  means  the  same  thing,  i.e.,  that  interest 
lags:  other  elements  of  the  sphere  of  money  and  credit  do  not. 


THE  RATE  OF  INTEREST 

On  the  whole,  this  is  what  we  find.  Owing  to  the  abnormal  sensitive- 
ness of  our  series  and  to  the  fact  that  short  money  rates  reflect,  and  are 
connected  with,  everything  that  happens  in  the  business  world  from  day 
to  day,  surface  movements  are  much  more,  and  underlying  movements 
correspondingly  less,  accentuated  than  in  a  series  of  employment  or 
pig-iron  consumption.  We  therefore  find  that  the  graph,  while  it  dis- 
plays many  other  surface  movements  also,  is  dominated  by  the  Kitchin 
wave,  presence  of  which  is  obvious  at  the  first  glance.  Interest  rate  is, 
in  fact,  the  standard  instance  by  which  to  demonstrate  this  cycle  and  has 
been  repeatedly  analyzed  with  a  view  to  measuring  it.  For  us  it  is 
sufficient  to  refer  again  to  Professor  Crum's  periodogram  analysis  of  the 
New  York  commercial  paper  rate  series  (Review  of  Economic  Statistics, 
1923)  which  has  been  several  times  quoted  in  this  book,  and  to  point  out 
that  the  phenomenon  can  readily  be  seen  also  in  the  English  and  the 
German  series.  Nevertheless,  as  has  been  emphasized  by  Professor  Crum 
and  as  is  proved  by  the  fact  alone  that  different  authors  have  invariably 
derived  somewhat  different  periods,  the  results  of  formal  analysis  are  not, 
in  themselves,  entirely  convincing.  They  cannot  be  expected  to  be  so 
with  material  that  is  subject  to  so  many  internal  and  external  irregulari- 
ties (see  Chap.  IV,  Sec.  D).  No  one  with  the  slightest  familiarity  with 
business  responses  can  ever  hope  to  "prove"  this  or  any  other  cycle  from 
time-series  evidence  alone,  since  this  would  require  that  periods,  ampli- 
tudes, and  behavior  in  phases  be  entirely  unaffected  by  any  external  fac- 
tors and  always  the  same  in  war  and  in  peace,  in  times  of  increasing  and  in 
times  of  decreasing  gold  production,  in  times  of  sound  money  and  in  times 
of  actual  or  threatened  disorders.  Still,  such  short  and  such  strong 
fluctuations,  which  also  are,  comparatively  speaking,  so  regular,  afford  an 
exceptionally  favorable  case.  We  cannot  hope  for  equally  good  results 
with  regard  to  any  much  longer  ones.  Formal  methods  would  hardly 
yield  results  in  the  case  of  the  Juglar,  and  if  formal  measurement  be 
insisted  on,  it  is  easy  to  deny  its  existence.  But  if  the  reader  follows  our 
series  year  by  year,  starting  with  the  characteristic  rise  in  1879,  he  will 
be  able  to  mark  off  the  last  two  Juglars  of  the  second,  and  the  first  two 
of  the  third  Kondratieff.  Some  well-marked  troughs  are  a  help — e.g., 
1885,  1894,  and  1904  (1908  requires  careful  interpretation) — but  those 
peaks  which  are  immediately  followed  by  abrupt  fall  of  over  3  per  cent 
should  be  neglected.  Of  course,  in  this  case,  as  in  all  others,  conviction 
can  be  attained  only  through  a  knowledge  of  what  actually  happened, 
strictly  speaking  from  year  to  year,  in  the  industrial  organism.  Chapters 
VI  and  VII  seem,  however,  to  provide  all  that  is  necessary  for  a  first 
attempt. 

In  the  series  of  commercial  paper  rates  and  its  equivalents  in  England 
and  Germany,  the  Kondratieff  should  show  in  what  traditional  analysis 


630  BUSINESS  CYCLES 

would  describe  as  breaks  in  trend — as  it  does  in  other  series  that  cover 
about  the  same  stretch  of  time.  Stating  the  question  is  answering  it. 
Everyone  knows  that  rates  displayed  a  long-time  tendency  to  fall  until 
the  middle  nineties,  when  an  opposite  tendency  began  to  assert  itself, 
although  in  the  United  States — not  so,  as  we  have  seen,  in  Germany 
and  England — bond  yield  failed  to  follow  suit  for  some  years.  It  is 
possible  to  draw  a  plausible  declining  "trend"  through  the  New  York 
figures  from  the  beginning  of  the  series  to  1897 — the  minimum  occurs  in 
1894 — and  a  not  less  plausible  rising  one  through  the  figures  for  1897  to 
1913.  The  same  can  be  done  for  England — where  the  minimum  in  mar- 
ket rate  occurs  in  1895 — and  Germany.  This  means  that  we  have  only 
a  slight  and  very  short  recovery  rise,  which,  however,  is  for  the  same 
reason  as  in  the  case  of  price  level  not  astonishing  in  a  Kondratieff,  the 
long  and  gentle  sweep  of  which  is  little  influenced  by  spirals.  We  may 
go  farther  back,  although  on  decreasingly  satisfactory  material.  The 
rise  in  yields  of  English  consols  toward  the  end  of  the  eighteenth  century 
does  not,  as  the  rise  in  price  level,  precede  the  French  wars,  but  coincides 
with  their  beginning.  The  all-time  peak  of  1798  could  be  sufficiently 
explained  by  those  wars  and  inflation  alone.  The  wars  also  explain  the 
fall  of  Prussian  governments,  which  yielded  over  16  per  cent  as  late  as 
1813.  This  casts  doubt  on  such  evidence  as  we  have  about  interest  rates 
from  1787  to  1815  at  least,  but  after  the  latter  year  we  are  entitled  to 
speak  of  a  tendency  of  interest  to  fall  for  a  period  much  too  long  to  be 
accounted  for  merely  by  the  process  of  absorption  of  war  effects,  some  of 
which  moreover  would  have  tended  to  raise  rates. 

That  period  lasts  in  England  and  Germany  to  between  1842  and 
1845.  For  this  country  disorders  of  the  wildcat  banking  type  and  cheap- 
money  policies  explain  violent  upward  spurts  in  the  thirties  (1834,  1836  to 
the  beginning  of  1842;  see  Cole  and  Smith,  op.  cit.,  table  on  pp.  192  and 
193)  which  blur  the  picture:  rise  indicative  of  the  processes  of  a  Kondra- 
tieff prosperity  here  begins  in  1845.  So  it  seems  to  have  been  also  in 
Germany,  and  this  is  a  more  serious  deviation  from  expectation,  since  it 
cannot  be  accounted  for  by  equally  significant  disorders  in  the  sphere  of 
credit  during  the  late  thirties.  For  1844  (see  Kahn,  op.  cit.,  and  Voye, 
op.  cit.)  bond  yield,  mortgage  interest,  and  the  rate  of  discount  of  the 
Prussian  Bank  were  all  at  a  minimum  of  respectively  3^,  4,  and  3j-£  per 
cent.  A  significant  rise  begins  in  1845,  which  clearly  links  up  with  the 
Kondratieff  processes  of  the  forties — railroads,  in  particular.  But  it 
lasts  until  about  1870,  when  yields  of  first-class  fixed  interest  securities 
were  about  5  per  cent.  We  might  question  the  evidence.  Requirements 
of  war  finance  and  the  speculative  excesses  of  the  fourth  Juglar  could  well 
be  held  responsible  for  what,  from  our  standpoint  certainly,  is  an  irregu- 
larity, and  there  were  significant  relapses.  Since  this  irregularity  is, 
however,  parallel  to  the  one  observed  in  the  behavior  of  price  level,  it  is 


THE  RATE  OF  INTEREST  631 

reasonable  to  connect  it  with  the  effect  on  demand  for  balances  that  a 
rising  price  level  would  have.  The  course  of  events  in  1870  was  domi- 
nated by  the  Franco-German  War,  and  during  the  next  three  years 
interest  rates  fluctuated  under  the  influence  of  the  mania  of  promotions1 
and  the  indemnity.  The  picture  is  somewhat  distorted  by  the  policy  of 
the  Prussian  Bank,  which  in  the  second  half  of  1872  resorted  to  restriction 
of  credit  rather  than  to  an  "adapted"  increase  in  rate.  After  the  crisis, 
both  short  rates  and  yields  return  to  form. 

English  rates  illustrate  our  mechanism  to  perfection.  We  will  glance 
at  the  market  rate  for  high-class  bills.  It  promptly  rises  at  the  beginning 
of  the  forties,  obviously  in  response  to  railway  building,  and  before  there 
was  any  rise  in  price  level.  The  year  1846  displays  the  Jugiar  peak  (the 
annual  average  for  1847  is  still  higher — 5  pounds,  17  shillings,  6  pence2 — 
but  this  was  a  panic  peak)  followed  by  the  lagged  decline  to  the  second 
half  of  1850  (average  for  the  year,  2  pounds,  5  shillings).  Then,  with  the 
second  Jugiar  rising  and  the  Kondratieff  prosperity  going  on,  the  rate 
again  rises — with  a  setback  in  1852  indicative  of  the  presence  of  a  counter- 
acting force,  gold  inflow — to  a  peak  higher  than  that  of  the  first  Jugiar  in 
1856  (5  pounds,  10  shillings,  the  figure  for  1857  is  again  higher,  6  pounds, 
15  shillings,  but  still  more  obviously  due  to  a  panic),  which  marks  the 
culmination  of  the  Kondratieff  very  well — no  such  annual  average 
occurring  again  within  the  period  except  in  the  financial  panics  of  1864 
and  1866.  Disregarding  these  and  also  the  intervening  year,  we  have  a 
falling  tendency  throughout  recession  and  depression  and  only  a  little 
upturn,  as  stated  above,  at  the  end  of  revival  (from  19  shillings,  2  pence 
in  1895  to  1  pound  18  shillings,  6  pence  in  1897).  Within  this  movement, 
Juglars  show  very  well — rise  in  1860  and  1861;  then  a  normal  fall  to 
1868,  interrupted  by  and  resumed  after  1863  to  1866;  rise  again  from 
1869  through  a  setback  to  the  Jugiar  peak  of  1872  (4  pounds,  1  shilling, 
9  pence,  1873  again  displaying  the  influence  of  the  panic  in  its  annual 
average  of  4  pounds,  14  shillings);  fall  to  1876;  recovery  rise  1877,  and 
some  panic  influence  in  1878;  after  reaction  to  this,  a  rise  in  1880,  1881, 
and  1882;  fall  to  1886;  no  substantial  change  until  the  lagged  rise  in  the 
prosperity  of  the  last  Jugiar  (in  1889);  a  Jugiar  peak  in  1890  (average 
3  pounds,  17  shillings,  7  pence);  then  the  fall  (some  panic  effect  in  1893, 
however)  and  the  recovery  mentioned  above.  Yield  of  consols  tells  in 
this  period  about  the  same  tale. 

The  reader  will  have  observed  that  some  figures  have  been  discarded 
as  panic  peaks,  others  as  setbacks,  a  procedure  which  of  course  stands  * 
and  falls  with  its  historical  warrant  for  each  case.     But  he  will  also 

1  We  recall  that  from  1871  to  1873  there  were  928  flotations  with  a  capital  of  2,781 
million  marks.  The  states,  however,  repaid  debts.  The  Reich  even  bought  fixed  interest 
securities  as  an  investment. 

*  See  Palgrave,  op.  dt.t  p.  33. 


632  BUSINESS  CYCLES 

observe  that,  if  due  attention  be  paid  to  the  interference  of  the  three 
cycles  with  each  other  and  to  the  influence  of  wars,  panics,  and  monetary 
disorders,  little  remains  to  be  accounted  for  by  factors  other  than  those 
that  make  up  our  cyclical  mechanism,  and  that,  in  particular,  the 
behavior  of  rates  within  each  phase  of  each  cycle  is  substantially  according 
to  expectation.  This  is  all  the  more  significant  because  of  the  variations 
in  gold  production  that  occurred  in  the  period.  As  a  matter  of  fact,  we 
have  seen  that  they  did  assert  themselves.  But  neither  their  short-run 
effect  in  depressing  rates  nor  their  long-time  effect  in  raising  them  was 
strong  enough  to  prevail  over  the  cyclical  movements.  No  country  was, 
taking  the  period  as  a  whole — Calif  ornian  gold,  of  course,  acted  primarily 
on  the  United  States — so  directly  exposed  to  the  impact  of  new  gold  as 
was  England.  Nowhere  can  actual  short-run  effects  be  so  easily  followed 
up.  Yet  interest  behaved  in  no  other  country  so  nearly  according  to 
theory  as  it  did  in  England. 

3.  It  is — for  prewar  times — not  possible,  but  it  is  fortunately  not 
necessary,  to  prove  fully  the  statistical  relation  between  interest  and 
profits,  which,  since  our  analysis  makes  the  latter  the  dominant  factor  in 
the  variations  of  the  former,  should  be  particularly  close  under  any 
normal  circumstances.  As  far  as  we  may,  however,  trust  the  testimony 
of  the  German  dividend  series  (see  Chart  XXXVIII),  it  is  in  fact,  very 
close:  American  figures  about  dividends  declared  give  the  same  impres- 
sion; so  do  railroad  earnings.  In  comparisons  with  other  series,  the 
extreme  emphasis  that  the  interest  rate  places  upon  short-time  surface 
fluctuations  in  business — this  is,  as  has  been  pointed  out,  the  only  differ- 
ence between  its  cyclical  behavior  and  that  of  employment;  see  Chart 
XX — and  result  trends  must  be  taken  into  account.  The  lag  in  interest 
rate,  although  it  does  not  always  show  significantly  and  although  it  is  in 
many  cases  difficult  to  identify  among  all  those  erratic  elements  which 
make  sequences  in  our  material  so  unreliable,  must  also  be  borne  in  mind. 
With  these  provisos,  we  can  say  that  interest  goes  well  with  pig-iron 
production  or  unfilled  orders  of  the  United  States  Steel  Corporation, 
and  that,  in  the  two  longer  cycles,  particularly  in  the  Kondratieff,  it 
displays  the  relation  to  total  output  which  we  expect  on  the  theory :  the 
long  periods  of  falling  interest  rate  are  also  the  periods  of  strongest 
increase  in  total  output.  The  relation  is  predominantly  positive  in  the 
Kitchins  as  they  appear,  if  a  least-square  trend  be  eliminated  from  the 
output  series.  The  short-run  relation  of  interest  to  building  is  incon- 
clusive. Interest  is  positively  correlated  with  outside  clearings,  outside 
loans,  and  outside  deposits  in  prosperity  and  depression,  also — though 
less  so — in  revival,  but  negatively  in  recession.  Since  the  latter  fact  is 
the  main  reason  why  there  is  a  trend  in  these  series,  trend  elimination  will 
again  produce  lagged  covariation  in  all  phases. 


THE  RATE  OF  INTEREST  633 

The  reader  will  have  observed  that  the  expectations  which  our  model 
yields  for  the  behavior  of  interest  are,  with  two  exceptions,  almost  exactly 
the  same  as  the  expectations  for  the  behavior  of  price  level.  One  of  those 
exceptions  is  the  lag,  which,  however,  is  important  only  for  short  fluctua- 
tions. The  other  is  the  downward  trend  in  prices.  But  barring  these 
there  should  be  all  but  exact  covariation,  because  even  external  disturb- 
ances and  the  internal  irregularities — these  with  the  exception  of  the 
panic  peaks  in  interest — are  in  most  cases  likely  to  affect  price  level  and 
interest  in  the  same  sense.  This  is  in  fact  what  we  find,  and  not  less 
in  the  long  than  in  the  short  run:  a  glance  on  the  pulse  charts  will  satisfy 
the  reader  that  this  relation  is  not  less  in  evidence  in  the  Kondratieff 
than  in  the  Kitchin  and  that  it  persists  in  the  most  varied  constellations 
of  external  factors.1  For  the  period  from  1886  to  1925  it  has  been 
the  subject  of  Mr.  Zinn's  most  interesting  study,  previously  quoted,2 
which  yields  the  conclusion  that  neither  variable  can  be  considered  the 
fundamental  "driving  force/*  but  that  both  "presumably  fluctuate  in 
response  to  an  unknown  common  set  of  generative  causes,  active  at  the 
root  of  the  system,"  and  that  "the  interest  rate  at  any  time  is  related  in  a 
systematic  manner  to  the  preceeding  values  of  wholesale  prices,  as  well 
as  to  the  concurrent  values." 

The  mere  fact,  however,  that  price  level  and  interest  rate  are  only 
two  of  very  many — or  of  several  aggregative — variables  suffices  to  prove 
that  the  deviations  from  strict  covariation  of  the  two  must  have  other 
than  random  causes.  In  particular,  the  strong  inverse  relation  that 
exists  between  the  short  fluctuations  of  deposits  devided  by  price  level 
and  the  short  fluctuations  of  rate  of  interest,  both  taken  with  reference 
to  their  Kondratieff  movements,  suggests  that  there  are  systematic 
reasons  for  such  imperfections  of  that  covariation  as  we  observe,3  and 
the  above  analysis  but  leads  to  the  door  of  more  intricate  problems. 

1  Of.  Mr.  Keynes'  Treatise  on  Money,  vol.  II,  Chap.  30  VIII:  the  Gibson  Paradox.     He 
rightly  calls  it  "  one  of  the  most  completely  established  facts  within  the  whole  field  of  quanti- 
tative economics  though  theoretical  economists  have  mostly  ignored  it"  which  "ought  to 
be  susceptible  of  some  explanation  of  a  general  character."     The  reason  why  such  explana- 
tion is  not  to  be  found  in  the  cyclical  covariation  of  prices  and  interest  is  (p.  201)  that  it 
is  a  long-period  "rather  than  a  strictly  short-period  phenomenon."     But  the  long-period 
covariation  envisaged  by  Mr.  Keynes  is  simply  the  covariation  in  the  Kondratieff  cycle. 
From  our  standpoint,  we  had  better  speak  of  a  Gibson  Katadox. 

2  Review  of  Economic  Statistics,  October  1927,  especially,  pp.  189-197.     That  period 
has  been  divided  into  three  subperiods— 1886-1899,  1899-1912,  1912-1925— accidentally 
covering  almost  exactly  different  Kondratieff  phases.    The  correlation  between  the  factors 
expressing  the  interest-price  relation  in  each  of  them  ("system  factors,"  as  Mr.  Zinn  sug- 
gestively calls  them)  is  as  high  as  0.82  for  the  first  subperiod  and  the  two  others,  and  0.90 
for  the  second  and  third  subperiods. 

8  That  idea,  the  result,  and  the  method  by  which  it  was  attained,  entirely  belong  to 
Dr.  C.  E.  Thomas. 


634  BUSINESS  CYCLES 

4.  Within  a  system  of  interdependent  quantities  there  is  no  point  in 
trying  to  label  some  as  determining  and  others  as  determined.  Within 
a  definite  process,  however,  which  runs  its  course  in  the  system  there  is 
point  in  asking  what  is  the  particular  role  of  every  element  in  the  sequence 
of  events.  In  this  sense  we  call  interest  consequential  because  it  is, 
within  our  process,  moved  away  from  its  neighborhood  value  by  the 
entrepreneurial  demand  for  balances  and  does  not  by  its  own  behavior 
cyclically  disrupt  that  neighborhood.  This  proposition  follows  from  the 
analysis  in  Chaps.  Ill  and  IV  and  cannot  be  disproved  by  a  post  hoc,  ergo 
propter  hoc  argument  exclusively  based  on  the  fact  that  prosperities  are 
invariably  preceded  by  moderate  rates  of  interest.  But  it  is  necessary 
to  safeguard  it  against  another  line  of  reasoning — one  that  we  have  met 
before  and  that  is  primarily  associated  with  Professor  von  Hayek's 
famous  book  Prices  and  Production.1  We  assume  perfect  equilibrium 
of  perfect  competition — coexisting,  however,  with  a  positive  rate  of 
interest2 — to  start  with. 

Then  there  will  be,  as  our  own  model  shows,  unused  facilities  for  the 
creation  of  balances.  Every  bank  has,  in  a  perfectly  competitive  member 
bank  system,  a  prima-facie  motive  to  make  use  of  them.  This  will  reduce 
the  rate  of  interest  below  its  neighborhood  value — the  real  or  natural  rate 
in  the  sense  of  that  group  of  authors — hence,  increase  the  value  of  all 
durable  goods,3  the  production  of  which  will  therefore,  absolutely  and 
relatively  to  that  of  transient  goods,  expand  in  response  to  this  stimulus 
and  turn  out  to  be  untenable  as  soon  as  the  latter  is  removed,  i.e.,  as  soon 
as  money  rate  again  equals  the  "real"  rate  of  interest — the  implication 
being  that  this  is  bound  to  happen  when  the  banks  run  up  against  the 
technical  limit  of  their  power  to  create  balances.  This  reasoning  seems 
to  give  a  rational  schema  of  cyclical  movements,  the  motive  force  for 
which  is  entirely  supplied  by  the  initiating  action  of  banks.  It  will  be 
shown  in  the  next  chapter  why  it  is  not  likely  that  banks,  even  if  they 
work  perfectly  competitively,  would  take  that  initiative  which  is  by  this 
theory  held  to  disrupt  the  existing  state'  of  equilibrium.  But  that  no 

1  We  are  here  interested  in  only  one  aspect  of  that  theory,  though  a  fundamental  one, 
and  we  do  not  intend  to  discuss  it  as  a  whole.     For  the  sake  of  argument  full  employment 
of  resources  at  the  outset  is  taken  for  granted.     On  other  points  see  Hansen  and  Tout, 
Annual  Survey  of  Business  Cycle  Theory,  Econometrica,  April  1933.     The  argument  we 
are  going  to  discuss  also  underlies  Professor  von  Mises'  and  Mr.  Hawtrey's  theories  of  the 
cycle  and  goes  back  to  Wicksell. 

2  This  assumption  is  also  granted  for  argument's  sake.    If  the  writer's  theory  of  interest 
be  accepted,  the  problem  we  are  about  to  discuss  does  not  arise  at  all. 

8  This  formulation,  somewhat  more  general  than  the  usual  one,  is  due  to  Professor 
Machlup,  who  has  also  in  a  very  telling  way  stressed  the  difference  in  the  effects  which 
variations  of  interest  exert  on  cost  of  current  production  and  cost  of  long-time  investment 
(of.  Interest  as  Cost  and  Capitalization  Factor,  American  Economic  Review,  September 
1935). 


THE  RATE  OF  INTEREST  635 

such  initiative  can  as  a  matter  of  fact  be  the  prime  mover  of  the  cyclical 
deviations  from  equilibrium  is  obvious  because,  though  interest  may  only 
rise  with  a  lag,  it  certainly  does  not  fall  in  prosperity,  or  at  its  beginning. 
This  would,  however,  be  necessary  in  order  to  produce  the  above  effect 
on  the  system  of  commodity  values.  It  cannot  be  replied  that  equilib- 
rium rate  must  be  defined  with  reference  to  the  demand  schedule  for 
balances  and  that,  hence,  if  that  schedule  shifts  upward,  as  it  does  in 
prosperity  right  from  the  start,  money  rate  may  be  "too  low"  or  even, 
relatively  to  the  data  of  the  situation,  "falling,"  even  if  it  actually  rises. 
For  first,  that  shift  would  take  precedence,  logically  as  well  as  in  time, 
over  any  initiative  of  banks — which,  in  fact,  would  as  an  element  of 
explanation  be  rendered  superfluous  by  it — and  would  have  to  be 
explained  independently,  either  by  the  action  of  external  factors  or  by 
innovation.  And  second,  for  old  firms  the  demand  schedule  for  balances 
is  shifted  only  by  the  effects  of  entrepreneurial  expenditure:  hence,  any 
expansion  of  their  productive  apparatus,  i.e.,  expansion  along  old  produc- 
tion functions,  could,  independently  of  that  expenditure,  occur  only  if 
interest  fell  absolutely,  which  it  does  not.  It  should  be  added  that  pro- 
duction along  the  new  production  functions  has  nothing  to  fear  from  a 
future  rise  in  interest,  since  it  is  protected  by  the  buffer  of  profit  and 
since,  moreover,  it  so  upsets  previous  conditions  that  it  is  by  no  means 
certain  that  time  preference — the  real  rate  in  the  sense  of  these  authors — 
will  remain  what  it  had  been.  It  is  much  more  likely  to  fall. 

But  if  we  thus  see  that  this  theory  fails  to  give  satisfaction  as  a  funda- 
mental explanation,  we  also  see  that  it  correctly  describes  a  large  expanse 
of  fact — as  has  been  pointed  out  repeatedly  in  the  course  of  our  historical 
sketch.  We  do  not,  of  course,  hold  that  the  behavior  of  banks  has 
nothing  to  do  with  the  cycle.  There  is  no  doubt  that  without  credit 
creation  amplitudes  of  cyclical  fluctuations  would  be  much  smaller, 
although  there  have  been  (witness  England  in  the  forties)  even  "manias" 
with  very  little  credit  creation,  and  although  the  effect  on  money  rates 
is  not  the  most  important  lever  through  which  it  works.  In  particular, 
the  phenomena  of  the  secondary  wave  would  then  be  much  less  in  evi- 
dence. Although  even  these  are  induced  not  by  money  rates  being  too 
low  but  by  entrepreneurial  activity,  higher  money  rates  would  go  far 
toward  keeping  them  in  bounds  and  low  money  rates  tend  to  foster  them. 
Judgment  in  granting  loans  is  much  more  important  than  the  rate 
charged,  and  reckless  banking  does  not  consist  in  financing  cheaply  but 
in  financing  irresponsibly.  But  effects  would  be  mitigated  if,  given  a 
certain  amount  of  irresponsibility,  credit  were  made  more  expensive. 

What  should  be  evident  more  than  anything  else  is  that  a  cheap 
money  policy  in  prosperity  can  have  no  other  effect  than  to  accentuate 
excesses  and  subsequent  breakdowns.  We  define  it  with  reference  to  the 


636  BUSINESS  CYCLES 

concept  of  Adapted  Rate  introduced  in  Sec.  A.  This  rate,  it  will  be 
remembered,  is  not  an  equilibrating  rate  and  does  not  prevent  the  system 
from  drawing  further  away  from  the  preceding  neighborhood,  although 
it  does  not  cause  its  excursion.  Cheap  money  policy,  being  the  attempt 
to  keep  actual  rates  below  it,  unavoidably  imparts  an  impulse  in  that 
direction.  A  dear  money  policy,  analogously  defined,  would  no  doubt 
also  be  effective  and  exert  equilibrating  influence.  But  in  neither  case 
can  effects  be  expected  to  be  as  great  as  the  usual  two- variable  analysis 
would  have  us  to  believe,  for  other  things  cannot  be  equal  under  the  cir- 
cumstances. Going  on  in  the  cyclical  sequence  we  notice  that  the  pri- 
mary factors  which  bring  about  the  upper  turning  point  are  independent 
of  the  rise  in  rates  that  has  previously  occurred.  In  this  sense  we  may 
say  that  interest  no  more  causes  the  down  turn  than  it  causes  the  excur- 
sion of  the  system  into  prosperity.  Navigation  becomes,  however,  diffi- 
cult beyond  this  proposition.  To  facilitate  it  we  will  at  first  disregard 
the  "lag"  and  assume  that  rates  move  synchronously  with  the  other 
relevant  elements  of  the  system.  Then  the  rate  will  move  pari  passu 
with  the  incipient  self-deflation  of  business.  But,  even  so,  it  will  press 
with  different  severity  on  different  sectors  of  the  system  and  certainly  be 
one  of  the  factors  to  make  the  operations  of  the  Secondary  Wave,  or 
some  of  them,  unprofitable.  And  cheap  money  policy  would  be  as  effec- 
tive to  counteract  this  as  action  on  any  single,  though  important,  element 
can  be.1  The  argument  is  at  first  sight  much  strengthened  by  the  fact, 
first  inserted  into  our  model  in  Chap.  IV,  that  important  lines  of  expendi- 
ture, some  types  of  dwelling-house  building,  in  particular — but  many 
adaptations  to  the  things  that  newly  emerged  in  the  preceding  prosperity 
and  a  great  part  of  the  conquest  of  the  new  economic  space  come  under 
the  same  heading — are  actually  highly  sensitive  to  variations  in  interest 
alone,  which  therefore  here  acquires,  as  any  other  element  can  more  or 
less,  a  "secondarily  causal"  role.  But  as  our  study  of  time  series  has 
shown,  interest  actually  does  fall,  by  virtue  of  the  working  of  the  cyclical 
mechanism,  fully  as  much  as,  and  more  than,  is  necessary  to  keep  up  sys- 
tem expenditure,  since  those  lines  of  expenditure  do,  in  fact,  expand 
in  recession.  The  case  for  accentuating  this  effect — in  practice  it  mostly 
rests  on  a  failure  to  distinguish  between  the  processes  of  recession  and 
the  processes  of  depression — would  therefore  have  to  rest  on  individual 
circumstances,  and  also  have  to  be  strong  enough  to  overcome  the  objec- 
tion to  keeping  alive  those  operations  of  the  Secondary  Wave  and  to 
preventing  the  liquidation  of  maladjustments.2 

1  In  formulating  as  above,  we  follow  the  fashion  in  assuming  that  the  effect  of  falling 
interest  on  the  consumers'  expenditure  of  the  strata  affected  is  negligible.    This  is,  however, 
not  necessarily  so. 

2  Lest  to  some  readers  this  should  seem  an  unpleasant  result,  the  writer  begs  to  point 
out  that  his  argument  can  be  made  to  imply  the  necessity  of  much  more  control,  regulation, 


THE  RATE  OF  INTEREST  637 

No  such  objections  to  cheap  money  arguments  apply  in  depression 
phases.  But  when  the  "crisis,"  if  any,  is  over,  interest  does  fall  as  a  rule 
very  considerably  and  sometimes  precipitately.  There  is  no  better 
example  by  which  to  demonstrate  how  little  the  rate  of  interest  can  by 
itself  do  than  the  typical  course  of  events  in  depressions.  With  more, 
though  still  little,  justification  could  it  be  argued  that  there  are  junctures 
at  which  its  increase  by  authoritative  action  would  be  more  effective  in 
stimulating  business,  because  it  would  call  forth  some  demand  for  bal- 
ances, which  holds  back  in  expectation  of  further  fall.  We  need  not  turn 
to  our  model  in  order  to  elaborate  the  point.  All  we  have  to  do  is  to 
invoke  common  business  experience  which,  but  for  the  phraseology  that 
has  developed,  would  be  enough  to  establish  the  truth  that  tampering 
with  the  rate  of  interest  in  depression  is  but  a  piece  of  political  liturgy. 
It  follows  that,  however  promptly  interest  falls,  it  would  not  on  that 
account  become  a  major  factor  in  bringing  about  the  lower  turning 
point  in  business.  This  is  independent  of,  but  much  strengthened  by,  the 
fact  that  revival  is,  at  least  at  first,  typically  a  revival  in  current  business 
but  not  in  investment,  and  in  the  cost  of  current  transaction  even  zero 
interest  would  as  a  rule  mean  much  less  than  a  moderate  reduction  in 
wages.  As  revival  wears  on,  however,  interest  resumes  the  "secondarily 
causal"  role  it  loses  in  depression,  and  cheap  money  policies  would 
become  effective  again.  But  as  we  may  infer  from  theoretical  consider- 
ations, and  as  we  can  see  statistically  and  historically,  interest  does  not 
and  cannot  rise  quickly  and  considerably  in  conditions  of  both  liquidity 
and  good  business  prospects.  As  far  as  it  does  rise,  this  rise  fills  an 
obvious  function :  the  effect  of  cheap  money  policy  can  in  these  conditions 
only  be  a  Hayek  effect. 

We  have  now  to  insert  the  Lag,  remembering  that  what  is  usually  thus 
referred  to  is  a  composite  of  many  different  elements,  some  of  which  are 
not  genuine  lags  at  all,  since  interest  holds  place  behind  other  elements 
in  the  cyclical  sequence.  The  cases  of  the  lag  in  short  rates  and  of  any 
lag  there  is  in  bond  yields  behind  short  rates  must  be  distinguished.  Very 
little  friction,  if  any,  enters  into  the  first.  But  any  failure  of  short  rates 
to  rise  promptly  in  prosperity,  however  caused,  must  tend  to  intensify  its 
primary  and  secondary  processes — especially  the  latter1 — and  induce  a 
great  many  operations  that  will  lead  into  difficulties  later.  A  lag  in  and 
after  the  upper  turning  point  has  the  effect  that,  with  business  deflating 

investigation  into  the  details  of  each  case  than  any  simple  recommendation  of  cheap 
money  policy,  hence  enough  of  public  activity  and  bureaucracy  to  gladden  their  hearts. 
See  next  Chap.  Sec.  A. 

1  It  should  be  observed  that  this  lag  is  also  due  to  the  institution  of  credit  creation:  if 
prosperity  were  exclusively  financed  by  savings,  rate  of  interest  would  not  only  rise  more 
but  also  more  promptly. 


638  BUSINESS  CYCLES 

itself,  a  moment  comes  in  which  the  adapted  rate  equals  the  equilibrating 
rate  and  after  which  it  acts  punitively,  in  strict  theory,  for  the  rest  of  the 
recession.  Though  this  effect  presupposes  and  does  not  cause  recession, 
it  yet  intensifies  it.  But  it  should  be  observed  that  we  have  not  the 
same  reason  to  expect  a  lag  in  the  turning  point  and  in  recession  as  we 
have  to  expect  one  in  prosperity.  If  such  lag  as  there  is  asserts  itself  in 
the  shorter  cycles — the  case  of  the  Kondratieff  is  doubtful — about  equally 
in  all  phases,  this  must  be  accounted  for  somewhat  differently  in  each  of 
them.  The  main  factor  in  recession  is  not  risk,  still  less  borrowing  in 
order  to  cover  deficits,  but  the  new  demand  which  replaces  that  of  the 
entrepreneurs,  partly  at  once  and  partly — later  on — when  the  rate  has 
already  begun  to  fall.  It  follows  that  that  punitive  effect  is  not  of  a 
nature  to  cause  general  disturbances,  although  it  does  increase  the  diffi- 
culties of  the  typical  Old  Firm.  In  particular,  no  such  general  punitive 
effect  should  be  inferred  from  a  lag  of  short  rates  as  against  prices.  For, 
as  we  know,  falling  prices  are  compatible  with  flourishing  business.  In 
depression  unemployed  balances  ride  their  attack  on  interest  rate,  and  any 
lag  in  this  phase  can  be  due  only  to  the  element  of  risk.  With  decreasing 
risk,  the  weight  of  those  balances  which  must  first  melt  before  the  rate  can 
recover  also  accounts  for  the  fact  that  is  particularly  in  evidence  during 
Kondratieff  prosperities  and  more  than  anything  else  responsible  for  so 
many  people's  unshakable  belief  in  the  efficacy  of  a  low  rate  of  interest, 
viz.,  that  the  lower  turning  point  of  the  shorter  cycles  so  frequently  occurs 
when  the  rate  of  interest  is  still  falling.  Since  it  is  business  activity  that 
pulls  it  up,  we  shall  not  wonder  at  it.  About  the  causal  importance  of 
this  favorable  factor  it  is  not  necessary  to  add  anything  to  what  has  been 
said  above. 

The  lag  in  bond  yields  as  against  short  rates  is  not  a  regular  phenom- 
enon at  all.  This  apart,  it  differs  from  the  lag  that  has  just  been  dis- 
cussed, in  that  friction  plays  a  considerable  part  in  it;  and  it  can  be  best 
observed,  on  the  one  hand,  in  the  upgrade  of  the  current  Kondratieff  in 
this  country  and,  on  the  other  hand,  frequently  in  deep  depression  after 
the  panic,  if  any,  has  passed.  But  the  first  instance,  like  others  that  could 
be  cited,  is  to  be  explained  on  special  grounds  that  have  nothing  to  do 
with  the  working  of  our  mechanism.  The  latter  instances  have  received 
more  than  their  due  share  of  attention.  These  cases  owe  their  existence 
to  the  fact  that  in  the  atmosphere  of  such  a  situation,  with  the  causation 
of  which  interest  has  little  to  do,  neither  borrowers  nor  lenders  care  to 
do  any  business  at  any  rate  whatsoever,  and  that  the  price  of  existing 
bonds  is  kept  down  by  the  knowledge  that  the  sources  which  service  them 
are  being  increasingly  impaired.  In  general,  short-time  covariation  of 
yields  and  short  rates  is  instantaneous. 


CHAPTER   XIII 


The  Central  Market  and  the  Stock 
Exchange 


A.  Banks  and  the  Pulse  of  Industry. — Turning  to  questions  concern- 
ing the  way  in  which  financial  facilities  are  provided  for  the  purposes  of 
providing  financial  facilities,  accommodation  for  providing  accommo- 
dation, we  move  still  further  away  from  the  motor  forces  of  our  process. 
In  doing  so,  we  shall  continue  to  use  the  general  schema  of  the  sphere  of 
banking  which  has  been  sketched  out  at  an  earlier  stage  as  an  instrument 
that  will  help  us  to  interpret  the  actual  working  of  the  banking  systems 
of  our  three  countries.  We  have  also  at  various  turns  of  our  way  had 
enough  glimpses  of  how  they  actually  function  to  enable  us  to  piece 
together  a  serviceable  picture.  Those  traits  which  are  most  relevant  to 
our  present  purpose  may  be  summed  up  and  somewhat  amplified  as 
follows. 

In  the  period  under  discussion  the  institutional  pattern  of  banking 
systems  (quality  of  their  personnel  and  their  traditions  included)  under- 
went considerable  changes,  even  since  Peel's  act  in  England,  the  found- 
ation of  the  Reichsbank  in  Germany,  and  the  establishment  of  the  National 
Banking  System  in  the  United  States.  But  most — though  not  all — of 
them  were  in  effect,  whatever  the  intention  may  have  been,  nothing  but 
adaptations  to  the  situations  created  by  our  process  and  in  fact  part  of 
the  latter.  Such  general  propositions  about  the  whole  period  as  we  are 
forced  to  make  should,  hence,  be  reformulated  for  each  of  the  historical 
subdivisions,  which  in  strictness  ought  to  be  dealt  with  individually. 

1.  The  individual  member  bank  may  be  so  big  as  to  be  able  to  influ- 
ence by  its  own  action  price  level  and  money  rates,  either  by  the  mechan- 
ical effect  of  its  operations  or  by  its  example.  The  member  bank  business 
of  the  Bank  of  England  is  an  outstanding  case  for  about  1845-1870.  In 
the  last  three  or  four  decades  of  our  period  the  great  English  and  German 
concerns,  toward  the  end  of  it  also  the  New  York  group  which  was  dubbed 
Money  Trust,  afford  other  instances.  Whether  a  bank  holds  such  a 
position  or  not,  however,  it  finds  itself  hedged  in,  not  only  by  familiar 
conditions  which,  though  elastic  in  the  long  run,  yet  technically  limit 

639 


640  BUSINESS  CYCLES 

the  sum  total  of  balances  it  can  put  at  the  disposal  of  its  customers  at 
any  given  point  of  time,1  but  also  by  the  fact  that  it  cannot  normally 
take  the  initiative  in  lending  or  even,  without  taking  the  initiative,  allow 
loans  mechanically  to  expand  to  that  limit.  This  simple  fact  is  so  much 
covered  by  the  embers  of  ancient  and  recent  controversy  and  so  liable 
to  being  misunderstood  that  we  must  explain  our  meaning  with  some 
care,  even  at  the  risk  of  repetition.  We  will  first  speak  of  member 
banks'  business  with  their  customers,  then  of  their  operations  in  the  open 
market,  and  finally,  of  a  particular  aspect  of  the  investment  item. 

When  nearly  a  century  ago,  in  the  controversy  on  the  principles  of 
banking  that  centered  around  Peel's  legislation,  Fullarton  held  the  view 
which  was  put  forth  again  and  again,  before  as  well  as  after,  by  writers 
representing  the  views  of  the  banking  world — namely,  that  banks  cannot 
"force  their  money  upon  people" — he  laid  himself  open  to  the  rather 
obvious  reply  that  they  have  a  strong  motive  (in  perfect  competition,  at 
least)  to  use  their  facilities  to  the  full  and  that,  hence,  they  will  so  frame 
their  conditions  as  to  call  forth  the  corresponding  demand.  This  reply 
fails  to  do  justice  to  what  Fullarton  really  meant  and  but  repeats  the  very 
error  that  he  probably  wished  to  point  out,  viz.,  the  error  implied  in  the 
application  to  banks  of  the  general  schema  of  business  behavior.  Any 
other  manufacturer  simply  wishes  to  sell  the  quantity  of  product  which 
will  yield  the  maximum  net  revenue  and  does  not  bother  about  what 
happens  to  the  units  after  he  has  sold  them.  The  manufacturer  of  bal- 
ances who  wants  "his  money"  back,  cannot  for  this  and  other  reasons 
behave  according  to  the  same  schema.  For  him  other  considerations 
enter  into  every  transaction  with  every  customer  and  make  of  it  an 
individual  case,  which  cannot  be  dealt  with  in  the  same  way  as  the  sale  of 
a  pair  of  boots.  Moreover,  every  one  of  these  individual  cases  is,  on  the 
one  hand,  an  element  of  his  relation  to  the  customer  which  must  be 
viewed  as  a  whole,  and  on  the  other  hand,  an  element  in  his  total  position 
which,  also,  must  be  watched  as  a  whole.  This  forces  upon  him  an  atti- 
tude of  resejrve,  which  is  entirely  absent  from  the  behavior  of  any  other 
businessman.  To  be  sure,  this  attitude  is  not  always  observed.  We 
have  met  examples  in  our  historical  report.  The  fact  that  often  it  is 
not  observed  is  the  very  reason  why  there  was  also  plenty  of  practical 
wisdom  in  the  teaching  of  the  so-called  currency  school  (Lord  Overstone) 
which  Fullarton  fought,  and  which  in  turn  fought  his  teaching,  according 
to  the  approved  method  of  economists,  which  consists  in  never  meeting 

1  This  distinction  between  the  limits  set  at  any  time  to  the  expansion  of  the  deposits 
of  all  the  member  banks  taken  together  and  to  the  expansion  of  the  deposits  of  any  single 
bank  is  part  of  traditional  teaching  and,  hence,  need  not  be  elaborated.  It  should,  how- 
ever, be  recalled  that  traditional  doctrine  stresses  inadequately,  first,  the  elasticity  of 
those  limits,  and  second,  the  fact  that  since  individual  banks  expand  and  contract  together, 
the  limits  set  to  the  expansion  of  an  individual  bank  lose  much  of  their  importance. 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    641 

one's  opponent's  point.  Nevertheless,  that  attitude  is  an  essential 
part  of  the  logic  of  a  banker's  situation.  And  it  follows  that,  while  most 
shopkeepers  will  normally  congratulate  themselves  whenever  they  are 
"sold  out,"  the  banker  does  not  typically  aim  at  being,  and  does  not 
congratulate  himself  if  he  is,  "loaned  up."  On  the  contrary,  this  means 
for  him  an  exceptional  and  undesirable  situation  of  embarrassment  and 
of  danger,  which  is,  as  a  matter  of  practice,  always  recognized  to  mean 
that,  both  by  the  individual  banker  himself  and  by  the  banking  com- 
munity. Customers'  business  cannot  be  handled  safely  and  comfortably 
unles^  each  bank  has  a  generous  allowance  of  unutilized  lending  power. 
Full  utilization  of  that  lending  power  in  member  banks*  business  with 
their  industrial  and  commercial  customers  is,  hence,  no  equilibrium  con- 
dition— nor  is  it  an  "adapted"  condition  outside  of  neighborhoods  of 
equilibrium — for  the  banking  system,  and  it  cannot  be  said  to  be  an 
obvious  interest  of  the  banker,  supplying  the  explanatory  principle  of 
his  behavior.  Economists  who,  on  the  strength  of  a  general  schema  of 
business  behavior,  insist  that  it  is  have  themselves  to  blame  for  the  quotes 
that  the  practical  banker  applies  to  their  "theories."  This  consideration 
completes,  from  the  practical  side,  an  argument  presented  in  the  preced- 
ing chapter,  and,  of  course,  applies  with  added  force  to  the  only  real 
case — that  of  imperfect  competition. 

Now,  initiative  may  mean  many  different  things.  It  is  not  suggested 
that  a  bank's  directing  committee  is  an  automaton.  Our  view  of  banking 
is,  in  fact,  much  less  mechanistic  than  that  of  those  theorists  who  attri- 
bute to  banks  the  role  of  prime  movers  in  business  cycles.  This  is 
obvious  from  the  emphasis  placed  throughout  our  analysis  on  the  element 
of  purpose.  For  us  it  is  neither  the  formal  character  of  the  business  to 
be  transacted — e.g.,  the  discounting  of  commodity  bills  of  exchange — 
nor  the  security  that  makes  sound  banking,  but  knowledge  and  under- 
standing of,  and  proper  attention  to,  the  purpose  which  the  balances 
applied  for  are  to  serve.  Judging  the  chances  of  success  of  each  purpose 
and,  as  a  means  to  this  end,  the  kind  of  man  the  borrower  is,  watching 
him  as  he  proceeds  and  granting  or  withholding  further  support  accord- 
ingly— these  are  the  fundamental  functions  of  that  committee  which  are 
more  important  than  the  mere  decision  how  far  the  bank  is  to  go  in 
granting  loans,  how  much,  if  at  all,  it  should  lean  on  the  bankers*  bank, 
how  great  a  risk  of  maneuvering  itself  into  a  tight  corner  it  should  under- 
take, and  so  on.  The  above  statement  that  a  bank  cannot  normally  take 
the  initiative  in  its  business  with  its  customers  merely  meant  that  it  can- 
not normally  initiate  the  individual  transaction.  Its  truth  becomes 
obvious  if  we  consider  what  that  would  involve.  As  far  as  the  financing 
of  enterprise  is  concerned,  it  would  involve  suggesting  definite  plans  to, 
or  urging  on,  people  who  have  every  motive  to  go  ahead  and  must  nor- 


642  BUSINESS  CYCLES 

mally  be  expected  to  know  the  ground  on  which  they  are  standing  better 
than  any  banker  can.  There  are,  no  doubt,  exceptions  to  this.  The 
practice  of  banks  of  the  credit  mobilier  type  supplies  the  most  important 
one,  especially  in  cases  like  mergers  or  in  cases  in  which  an  innovation 
acquires  additional  importance  from  simultaneous  developments  not 
within  the  horizon  or  radius  of  influence  of  the  man  who  is  to  carry  it  out. 
Independently  of  this  there  are  occasions — they  may  be  the  high  points  in 
the  career  of  a  great  banker — in  which  a  bank  can  successfully  make  itself 
responsible  for  an  enterprise  by  pledging  its  support  and  committing 
itself  to  seeing  the  entrepreneur  through.  But  it  is  evident  how  risky 
this  is.  As  a  rule,  it  augurs  ill  for  a  proposition  if  it  has  to  be  forced  on 
the  primarily  responsible  man.  But  the  same  applies,  in  a  lesser  degree, 
even  to  current  transactions.  A  bank  often  sees  reason  to  restrain,  but  it 
is  rarely  in  a  position  to  ask  its  customer:  Won't  you  borrow  in  order  to 
do  this  or  that  ?  The  chief  exception  with  some  banks  in  some  countries 
is  the  role  they  assume  in  dealing  with  private  investors  or  speculators. 
But  that  is  another  matter. 

It  might  be  urged  that  initiative  need  not  go  as  far  as  this  but  can  be 
exerted  without  any  particular  suggestions  by  assuming  a  general  attitude 
of  encouragement,  which  consists  mainly,  though  not  wholly,  in  offering 
attractive  conditions  and  in  conveying  to  wavering  customers  the  impres- 
sion that  if  they  go  ahead  they  will  not  do  so  alone.  Even  a  small  bank 
may  do  that  if  it  observes  that  others  do  it,  but  a  bank  important  enough 
to  influence  situations  by  its  own  action  or  example  seems  particularly 
able  to  impart  such  a  stimulus.  Hence  the  question,  which  in  fact  is  very 
frequently  asked,  why  the  banking  system  does  not  use  this  power  or  even, 
as  some  would  put  it,  why  it  uses  it  viciously,  i.e.,  so  as  to  intensify  both 
booms  and  depressions.  Recalling  our  discussions  of  historical  cases 
and  the  relevant  argument  in  the  preceding  chapter,  we  may  first  repeat 
that  no  encouragement  from  banks  is  necessary  in  order  to  start  a  pros- 
perity phase  and  that  such  encouragement  as  has  been  actually  proffered 
— it  mostly  comes  within  our  concept  of  reckless  banking,  witness,  for 
example,  the  events  preceding  1837  and  1907 — can  be  shown  to  bear  a 
close  relation  to  the  occurrence  of  crises  and  downward  spirals;  second, 
that  no  discouragement  from  banks  is  necessary  to  bring  about  the  turn- 
ing point  into  recession  in  our  sense,  and  that  such  discouraging  influence 
as  is  sometimes  exerted  only  serves  at  that  juncture  to  steady  and  not  to 
dislocate  the  system;  and  third,  that  near  or  at  the  lower  turning  point 
interest  rate  is  in  any  case  low  and  falling  and  bankers'  attitude  as  a  rule 
encouraging  enough.  That  question,  therefore,  narrows  down  to  the 
behavior  of  banks  before  and  during  depression.  It  has  been  pointed  out 
in  the  preceding  chapter  that  the  end  of  recession  is  the  only  stage  to 
offer  possibilities  of  corporative  initiative  by  banks.  The  reason  why 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    648 

more  use  is  not  made  of  these  possibilities  is  analogous  to  the  reason  why 
initiative  action  by  banks  could  do  but  little  once  a  downward  cumula- 
tive process  has  set  in.  Banks  control  one  element  only  of  the  situation 
in  which  businessmen  find  themselves,  and  that  situation  contains  so 
many  hitches  and  untenable  positions  that  such  action,  besides  becom- 
ing extremely  difficult,  cannot  be  expected  to  be  effective.  But  it  is 
true  that  the  survival  interest  of  each  individual  bank  then  drives  it  into 
courses  of  action  which,  even  without  panicky  calling  of  loans — very 
important  though  that  element  is  in  the  mechanism  of  crises,  and  though 
it  was  especially  so  during  the  earlier  part  of  the  prewar  period — tend  to 
intensify  the  phenomena  of  the  spiral.  All  this  only  amounts  to  recog- 
nizing again,  this  time  from  the  side  of  banking  practice,  that  interest  is 
not  the  hero  of  the  cyclical  drama. 

No  apologetic  purpose  is  either  intended  or  served  by  this  analysis. 
On  the  contrary,  its  suggestions  could,  if  this  were  within  our  purpose,  be 
worked  up  into  a  very  comprehensive  plan  of  regulation;  but  it  would  have 
to  be  primarily  restrictive  in  nature,  although  not  wholly  so.  And  it 
would  have  to  aim  at  improving  personnel  and  at  enforcing  adherence  to 
standard  practice  rather  than  different  principles  of  practice.  Machinery 
which  would  enable  some  authority  to  force  banks  to  take  initiative  action 
would,  if  effective  at  all,  in  most  cases  lead  to  additional  maladjustments. 
In  those  cases,  moreover,  in  which  such  machinery  could  be  expected  to 
have  remedial  or  preventive  effects  it  would  still  be  inferior  to  a  policy 
that,  while  leaving  the  banks  free  to  fill  their  function,  would  directly 
act  on  the  economic  process — by  government  expenditure,  for  example. 

2.  To  economists  who  can  see  in  the  business  cycle  nothing  but  an 
effect  of  the  working  of  the  banking  system,  and  particularly  in  depressive 
situations  nothing  but  "deflation,"  which  could  almost  at  will  be  turned 
into  prosperity  by  "  refl  ationary  "  creation  of  deposits  defeating  the 
attempt  of  business  to  deflate  itself,  our  argument  may  well  be  inaccept- 
able,  although  it  seems  to  agree  with  rather  obvious  facts  of  business 
experience — which  is  the  reason  why  we  fared  so  well,  in  discussing  the 
behavior  of  clearings,  deposits,  and  interest,  with  the  working  hypothesis 
of  a  "passive"  banking  system.1  Nor  will  those  agree  for  whom  pros- 
perity and  depression  differ  in  nothing  but  the  state  of  businessmen's 
minds — elated  in  the  one  condition  and  depressed  in  the  other.  For  if 
there  were  nothing  in  the  objective  situations  to  account  for  those  humors, 
it  would  be  plausible  to  assume  that  psychotherapeutic  behavior  on  the 
part  of  banks  would  never  fail  to  be  effective.  However  strong  the 

1  It  should  be  observed,  however,  that  the  word  passive  does  not  express  our  meaning 
well  and  might  easily  prove  misleading.  The  preceding  paragraphs  sufficiently  show  that 
we  do  not  mean  more  than  that  banks  primarily  act  in  response  to  business  situations, 
which  cannot  in  turn  be  explained  by  their  action. 


644  BUSINESS  CYCLES 

evidence  which  verifies  our  view,  it  cannot  be  made  fully  convincing 
except  by  reference  to  the  whole  of  our  analysis  of  the  cyclical  process. 
But  some  points  still  remain  that  require  specific  notice.  They  all  link 
up  with  member  banks'  operations  in  the  open  market. 

It  has  been  stated  above  that  while  the  behavior  of  banks  during  or  in 
expectation  of  breakdowns  may  acquire  causal  significance  for  some  of  the 
features  of  "crises,"  the  upper  turning  point  is  not  normally  brought 
about  by  banks'  calling  loans  because,  having  run  up  "against  technical 
limits,  they  are  losing  cash  and,  therefore,  have  to  retrace  their  steps. 
"The  greater  number  of  cyclical  fluctuations  keep  well  within  these 
limits,"  even  in  this  country,  and  there  is  "no  basis  for  the  belief  that 
these  cyclical  swings  never  halted  until  the  resources  of  the  banks  had 
been  exhausted."1  But  banking  systems  seem  to  approach  those  limits 
much  more  closely  than  they  actually  do,  because  they  do  not  keep  the 
whole  of  their  surplus  funds  in  the  form  of  cash  and  balances  with  other 
banks  but  allow  part  of  them  to  earn  interest  by  means  of  temporary 
investment.  In  substantial  agreement  with  practice,  we  call  such 
investment  the  Secondary  Reserve.  Now  this  policy  is,  indeed,  what 
the  customers'  business  is  not,  viz.,  first,  actuated  exclusively  by  the 
desire  to  use  funds  as  much  as  possible;  second,  entirely  a  matter  of  the 
initiative  of  the  banks'  managing  committees;  and  third,  as  mechanical  or 
routine  as  anything  in  business  can  be.  As  soon  as  committees  have 
settled  what  credits  they  are  to  grant  to  their  customers,  the  surplus  to  be 
temporarily  invested,  in  all  normal  situations  and  excepting  cases  of 
speculative  booms  that  make  the  open  market  abnormally  attractive,  is 
determined  and  will  be  invested — as  far  as  member  banks  are  concerned — 
with  little  regard  to  the  general  business  situation,  which  asserts  itself 
more  in  the  kind  of  investment  to  choose  than  with  reference  to  the 
decision  to  invest. 

This  is,  as  we  have  seen  in  Chap.  XII,  what  pulls  down  open-market 
rates  so  drastically  in  depression  as  soon  as  panics,  if  any,  are  over,  and 
even  in  the  other  phases  normally  tends  to  keep  them  at  a  low  level.  As 
we  have  also  seen,  effects  extend  to  the  stock  exchange  and  in  particular 
to  bond  prices.  But  there  they  stop  until  business  reacts  to  more  funda- 
mental stimuli.  Even  bond  issues  do  not  react  to  low  open-market  rates 
alone.  Conversely,  progressive  liquidation  of  temporary  investments 
when  business  revives,  restores  normal  financial  habits  and  helps  to 
raise  rates  again.  However,  even  if  banks  were  always  "loaned  up"  in 
this  sense,  the  same  consequences  would  not  follow  as  if  they  were  loaned 
up  in  the  relevant  sense,  i.e.,  in  their  customers'  business.  In  particular, 
they  need  not  normally,  if  "loaned  up"  in  the  wider  sense,  restrict  the 
latter  on  the  ground  that  they  have  expanded  to  a  technical  limit — 
1  A.  A.  Young,  op.  cit.,  p.  28. 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    645 

although  as  prosperity  wears  on  other  reasons  may  suggest  themselves  for 
doing  so — and  the  very  presence  of  secondary  reserves  outside  the  depres- 
sion phase  is  proof  that  they  never  want  to  be  "loaned  up"  in  the  narrow 
sense. 

What  secondary  reserves  consist  in  is  a  question  of  fact  to  be  answered 
differently  for  different  times  and  countries.  Where  there  are  facilities 
for  rediscounting,  high-grade  bills  of  short  usance,  short -time  government 
paper,  and  so  on  are  the  classical  items;  where  such  facilities  are  lacking, 
call  loans  to  the  stock  exchange.  It  depends  on  the  organization  of 
markets  and  on  the  traditional  behavior  of  the  public  what  other  "embod- 
iments of  future  balances"  qualify  for  the  role.  Government  bonds, 
possibly  also  other  bonds, may  so  qualify;  and,  where  banks  have  the  legal 
power  of  holding  it  and  trading  in  it,  even  common  stock.  In  the  case 
of  German  banks,  the  motive  of  employing  surplus  funds  in  this  way 
undistinguishably  merges  with  the  motive  of  supporting  the  prices  of 
securities  of  the  concerns  they  patronize.  It  should  be  observed  that 
each  member  bank  thus  fulfills  with  reference  to  firms  and  households  and 
within  its  sphere  of  influence  a  function  akin  to  that  which  bankers' 
banks'  open-market  operations  fulfill  with  reference  to  member  banks. 
It  buys,  and  thereby  supplies  the  public  with  liquid  funds,  on  the  down- 
grade; it  sells,  and  thereby  curtails  the  public's  available  means — a  qualifi- 
cation will  presently  be  noticed — on  the  upgrade.  This  mechanism  does 
not  work  alone  or  in  a  world  in  which  other  things  are  equal  and  therefore 
its  regulative  effects  are,  no  doubt,  often  overshadowed  by  many  other 
factors.  But  the  profit  motive  is  sufficient  to  set  it  into  motion. 
In  the  case  of  common  stock,  this  is  obvious.  In  the  case  of  bonds,  the 
practice  under  discussion  means  that  banks  buy  when  bonds  are  dear  and 
sell  when  their  prices  have  fallen — accentuating  the  fluctuations  in  bond 
prices  which  would  otherwise  prevail.  Such  operations  are  remunerative, 
nevertheless,  since  the  funds  engaged  in  them  would  otherwise  be  idle. 
So  we  see,  again,  that  neither  machinery  nor  motive  is  absent  for  meeting 
any  pessimistic  moods  of  the  public  by  the  creation  of  balances.  Only, 
there  is  little  point  in  arguing  that  it  should  be  made  more  effective,  since, 
as  pointed  out  in  Chap.  XI,  even  the  balances  actually  created  tend  to 
become  as  idle  as  the  corresponding  funds  of  the  banks  would  be. 

There  can  be  little  doubt  that  the  investment  item  of  the  American 
banks  was  for  our  period  mainly  of  the  nature  of  secondary  reserve. 
There  was  a  significant  special  trend  in  it.  But  both  New  York  and 
outside  banks — which  do  not  differ  as  much  in  their  investments  as  they 
do  in  loans — purchased  bonds  primarily  when  their  surplus  funds 
increased  and  sold — or,  at  all  events,  purchased  less — when  their  cus- 
tomers' business  expanded.  In  the  case  of  outside  banks,  though  not  of 
those  in  New  York,  this  meant  strong  inverse  relation  in  the  fluctuations 


646  BUSINESS  CYCLES 

in  investments  to  the  fluctuations  in  loans  and  discounts,  which  suffices 
to  establish  our  point.1 

3.  Of  course,  member  banks'  investments  cannot  entirely  be  accounted 
for  by  the  secondary -reserve  hypothesis.  To  begin  with,  political  neces- 
sity or  pressure  may,  in  times  of  abnormal  government  expenditure, 
compel  banks  to  increase  their  investment  in  government  bonds  to  an 
extent  altogether  beyond  the  range  of  secondary-reserve  considerations. 
Second,  we  know  that  one  method  of  financing  real  investment,  or  even 
current  expansion,  is  to  sell  assets  to  banks  (Chap.  XI),  an  operation 
which  differs  from  borrowing  in  technique  only,  but  tends  to  impart  to 
the  investment  item  a  time  shape  exactly  opposite  to  the  one  we  have 
described.  The  statistical  picture  proves  that  for  the  prewar  period  the 
latter  prevailed.  This  need  not  always  be  so,  however.  During  and 
after  a  time  of  abnormally  great  government  expenditure,  for  instance, 
not  only  banks  but  the  public  may  be  so  saturated  with  government 
bonds  that  buying  and  selling  them  may  well  become  the  central  element 
of  their  dealings  with  each  other.  In  this  case,  firms  and  households 
would,  on  the  one  hand,  be  much  more  independent  of  normal  bank 
credit  and,  in  consequence,  of  the  advice  and  the  approval  of  their  banks, 
and  it  would  be  much  easier  for  them,  or  some  of  them,  to  resort  directly 
to  the  open  market  in  order  to  finance  themselves.  On  the  other  hand, 
the  banks,  while  losing  their  hold  on  industrial  operations,  would  be 
much  more  in  a  position  to  act  initiatively  within  the  sphere  left  to  them 
under  such  circumstances.  They  could  then  with  more  justice  be  said  to 
"regulate  the  flow  of  funds"  and  to  control  the  volume  of  balances. 
The  postwar  period  illustrates  this.2  Only  it  would  not,  on  that  account, 
be  any  truer  to  say  that  they  thereby  regulate  the  pulse  of  business. 
On  the  contrary,  the  momentous  change  which  permanence  of  such 
conditions  would  imply — and  developments  since  1930  strongly  suggest 
that  these  conditions,  and  in  particular  government  expenditure  that  is 
abnormal  from  the  standpoint  of  capitalist  logic,  may  have  become 
permanent  by  now — precisely  consists  in  paralyzing  such  regulative 
influence  as  they  had  beyond  mere  regulation  of  the  amount  of  balances 
which  may  or  may  not  be  used.  An  essential  piece  of  capitalist  mecha- 
nisms would  be  gone  forever.  It  is  easy  to  see  that  it  would  be 

1  The  behavior  of  notes  in  circulation  accords  perfectly  with  our  view.     Barring  the 
influence  of  the  introduction  of  the  2  per  cent  consols  by  the  Act  of  1900,  they  fluctuate 
much  as  money  outside  of  treasury  and  banks  and  with  the  rate  of  interest.     They  thus 
followed  the  pulse  of  customers'  business,  as  we  should  expect. 

2  In  spite  of  the  sentence  that  follows  in  the  text,  it  should  be  pointed  out  that,  since 
most  adherents  of  what  we  have  called  the  Investment  Theory  of  Banking  (Chap.  Ill) 
as  a  matter  of  fact  mainly  reason  on  postwar  facts  and  problems,  the  difference  between 
our  view  of  banking  and  those  of  most  economists  of  our  day  is  not  quite  so  great  as  it 
seems. 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    647 

government,  acting  not  through  the  mere  offer  of  facilities  but  through 
expenditure,  and  not  the  banks,  which  might  then  really  acquire  something 
of  the  power  that  is  being  wrongly,  though  usually,  attributed  to  the  banks : 
these  would  be  more  powerless  than  ever  in  starting  the  movement  of  the 
economic  body  from  troughs  to  higher  levels  of  activity,  though  they 
would  have  the  same  power  they  always  had  to  foster  excesses. 

Third,  the  investment  item  transcends  secondary-reserve  considera- 
tions and  acquires  additional  importance  also  in  the  case  of  banks  which 
on  principle  participate  in  the  ventures  they  finance — roughly,  in  the 
case  of  the  crSdit  mobilier  type,1  as  exemplified  by  German  banks.  It  is 
part  of  the  task  of  any  bank's  managing  committee  to  exert  what,  in 
distinction  from  business  initiative,  we  may  term  financial  initiative, 
i.e.,  to  suggest  to,  sometimes  to  impose  on,  its  industrial  customers 
financial  policies  and  in  particular  answers  to  the  questions  how  far  and 
how  long  they  are  to  finance  themselves  by  borrowing  in  current  account 
and  when  and  to  what  extent  they  are  to  fund  these  debts  by  the  issue  of 
bonds  or  stock. 

As  we  have  seen  (Chap.  XI),  such  refunding  either  annihilates 
deposits  or,  if  the  bank  (or  another  bank)  finances  subscriptions,  sub- 
stitutes new  debtors  for  the  old  one,  and  in  both  cases  powerfully  affects 
not  only  the  situation  of  the  bank  itself  but  also  the  whole  banking  system 
and  the  open  market.  Issues  intended  to  finance  real  investment  may 
be  said  to  do  the  same  thing  potentially.  But  banks  of  the  mobilier 
type  themselves  acquire  parcels  of  stock,  either  because  this  strengthens 
their  hold  on  the  current  business  of  customers  and  their  claim  to  manage 
the  future  issues  of  the  same  concerns,  or  because  they  wish  to  influence 
the  latter's  business  policies — particularly  to  acquire  positions  of  strategic 
value  in  negotiating  amalgamations — or  because  a  given  issue  is  not  yet 
ripe  for  introduction  to  the  public,  or  because  they  intend  to  trade  in  it 
permanently,  or  simply  in  order  to  profit  by  appreciation.  Such  invest- 
ments, which  in  many  cases  develop  into  the  most  valuable  part  of 
the  assets  of  a  bank  and  into  the  backbone  of  its  business,  and  which  enter 
the  balance  sheet  either  as  "permanent  participations"  or  as  "securities" 
(Effekteri),  are,  of  course,  built  up  for  their  own  sake  and  are  largely, 
though  not  wholly,  exempt  from  the  considerations  incident  to  what 
English  or  American  opinion  holds  to  be  normal  banking  business. 
The  statistical  implications  of  this  need  no  emphasis.  Other  aspects  have 
been  discussed  in  Chap.  VII.  As  has  been  pointed  out  there,  even  in  this 
case  the  real  influence  of  banks  does  not,  in  general,  amount  to  control 

1  It  should  be  borne  in  mind  that  the  above  expression  is  used  because  the  writer  sup- 
poses that  it  conveys  what  he  means.  We  know  that  historically  it  is  not  quite  correct. 
See  Chap.  VII. 


648  BUSINESS  CYCLES 

of  what  is  being  done,  although  it  often  amounts  to  control  of  the  share- 
holders' meeting. 

B.  The  Central  Market  (in  an  Isolated  Domain). — First,  the  limited 
purpose  of  the  following  discussion  of  the  role  of  banker's  banks  in  the 
cyclical  process  must  be  borne  in  mind  throughout.  A  very  misleading 
impression  would  be  gathered  by  the  reader  if  from  the  points  to  be  touched 
he  tried  to  construct  a  general  theory  of  central  banking.  Second, 
complete  analysis  would  have  to  consist  of  a  string  of  historical  inter- 
pretations of  individual  central  banks  or  central  bank  systems,  some  frag- 
ments of  which  have  been  offered  in  previous  chapters.  Beyond  a  small 
number  of  propositions  that  may  be  true  of  every  institution  or  class  of 
institutions  that  can  be  partly  or  wholly  identified  with  central  banking, 
it  is  extremely  dangerous  to  generalize,  because  environments,  organiza- 
tions, habits  of  doing  business,  attitudes  of  and  toward  member  banks, 
firms,  and  speculation,  relations  to  government,  and  so  on  differ  so  widely 
as  to  impart  different  meaning,  at  different  times  and  in  different  coun- 
tries, even  to  identical  forms  and  phraseologies.  Two  instances  will  be 
sufficient  to  emphasize  this  once  more.  A  universal  spirit  of  sound 
money  and  sound  banking  in  the  classical  sense  pervaded  the  English 
business  and  banking  communities  for  most  of  our  period.  This  made  all 
the  difference  for  the  nature  and  effects  of  any  measures  taken  by  the 
Bank  of  England:  exactly  the  same  measures — and,  of  course,  both 
policies  and  phraseologies  of  English  banking  were  widely  copied — taken 
by  a  bankers'  bank  in  a  country  of  different  mentality  would  have 
differed  from  the  paradigma,  both  in  economic  meaning  and  in  effects. 
Again,  we  shall  later  notice  the  importance  for  the  policy  of  the  Bank  of 
England  and  for  the  results  it  produced,  of  England's  creditor  position. 
For  the  moment  we  defer  consideration  of  gold  movements,  international 
relations,  and  exchanges,  and  consider  central  banks  as  if  they  functioned 
in  isolated  domains.  But  even  so,  the  difference  of  which  that  creditor 
position  was  one  of  the  consequences,  remains  essential,  viz.,  presence  at 
any  given  point  of  time  throughout  the  period  of  a  relatively  great  amount 
of  accumulated  wealth  which  behaved  according  to  an  established 
tradition  and  made  things  very  much  easier  for  the  central  institution 
than  they  were  in  other  countries.  This  fact  alone  suffices  to  make  the 
application  of  the  principles  of  English  banking  to  other  countries  as 
doubtful  a  matter  as  the  application  of  English  principles  of  parlia- 
mentary government  to  other  national  patterns  has  been  or  would  have 
been. 

Third,  our  central  market,  consisting  of  the  transactions  between 
the  bankers'  bank  and  member  banks  and  the  operations  of  the  former 
in  the  open  market,  is,  of  course,  a  bold  abstraction.  From  previous 
discussions  it  should  be  obvious  that  in  the  case  of  the  Bank  of  England, 
for  instance,  our  concept  of  transactions  between  it  and  member  banks  is 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    649 

(in  part)  a  useful  fiction  representing  actual  fact.  But  it  is  (in  part)  a 
fiction  all  the  same.  Moreover,  until  the  foundation  of  the  Federal 
Reserve  System,  hardly  any  central  bank  confined  itself  to  bankers' 
bank  business  m  our  sense,  and  no  central  bank — not  excepting  the 
Federal  Reserve  System — ever  covered  all  the  functions  we  attribute  to 
bankers'  banks.  As  regards  the  latter  point,  we  ought  really  to  speak, 
even  in  the  English  case,  of  a  bankers'  bank  system  rather  than  of  a  single 
Central  Bank.  As  regards  the  former  point,  the  historical  importance 
of  central  banks'  member  bank  business  has  been  emphasized  before. 
It  should  be  added  that  this  business  was  a  most  important  lever  for 
influencing  the  structure  of  credit  and  greatly  strengthened  the  hands 
of  the  central  bank  in  its  bankers'  bank  functions:  it  gave  it  a  hold  on  the 
market  which  any  mere  bankers*  bank  must  always  find  it  difficult  to 
acquire.  That  in  some  respects  it  also  fettered  the  hand  it  made  stronger 
will,  on  reflection,  be  seen  not  to  involve  any  contradiction. 

Finally,  in  the  American  case,  some  readers  have  probably  taken 
offense  before  this  at  our  associating  central -bank  functions  in  this  country 
primarily  with  New  York  banks.  This  seems  to  run  counter  to  obvious 
facts,  the  outstanding  one  being  that  under  the  rule  of  the  National 
Banking  System  as  amended  in  1864  and  1887  a  system  of  reserve  and 
central  reserve  city  banks  developed,  which  legislation  itself,  by  imposing 
special  reserve  requirements — in  the  case  of  central  reserve  city  banks, 
also,  by  insisting  on  reserves'  being  kept  entirely  in  their  vaults — recog- 
nized as  bankers'  banks  and  balances  with  which  other  banks  were  allowed 
to  count  as  part  of  their  legal  reserves  to  %  of  the  total  amount  of  the 
latter.  This  practice  of  using  checking  accounts  with  certain  banks  as 
reserves,  or  of  redepositing  reserves  and,  generally,  surplus  funds  with 
correspondents  in  more  central  positions,  of  course,  established  a  relation 
of  current  cooperation  between  the  latter  and  their  banking  customers 
(brokers,  bond  houses,  investment  trusts,  and  so  on  as  usually  included). 
Correspondents  undertook  to  act  as  agents  for  these  as  for  other  cus- 
tomers, advised  them,  did  their  foreign-exchange  and  open-market  business 
for  them,  used  them  in  turn  as  agents  for  collection  and  other  pur- 
poses, and,  last  but  not  least,  accommodated  them  in  cases  of  seasonal  and 
cyclical  tension.  This  bankers'  bank  business  was  highly  competitive,  and, 
hence,  worked  out  very  advantageously  for  the  member  banks  in  our 
sense  of  this  term.  They  received,  not  only  a  considerable  amount  of  gra- 
tis service,  but  also  interest  on  some  rough  average  of  their  credit  balance, 
which  for  the  two  last  decades  before  the  Federal  Reserve  Act  is  usually 
estimated  at  2  per  cent.1  According  to  the  Comptroller's  reports,  about 

1  On  the  other  hand,  the  correspondents,  unless  they  confined  their  business  with  non- 
bank  customers  to  local  proportions,  were  hardly  able  to  go  without  these  connections. 
This  fact,  the  not  very  profitable  character  of  these  connections  in  themselves,  and  the 


650  BUSINESS  CYCLES 

half  of  the  deposits  of  the  banks  in  central  reserve  cities  were  bankers' 
balances.  Under  these  circumstances,  our  reliance  on  New  York  bank 
figures,  of  course,  means  relying  on  a  sample  of  American  bankers'  bank 
activities.  But  this  sample  includes  the  roof  of  the  structure  and  its  most 
important  elements,  and  as  has  been  mentioned  before,  the  behavior  of  the 
figures  goes  far  to  justify  our  choice. 

We  start  from  the  analogy  that  we  have  previously  noticed  between 
the  relation  of  a  member  bank  to  its  customers,  firms  and  households, 
and  the  relation  between  a  bankers'  bank  and  the  member  banks  it  banks 
for.  Once  the  nature  of  the  variations  in  system  expenditure  in  the 
cyclical  process  of  evolution  and  their  relation  to  the  balance-manu- 
facturing activity  of  member  banks  (member  creation,  as  we  call  it)  is 
fully  understood,  the  fundamental  questions  of  principle  regarding  the 
balance-manufacturing  activities  of  bankers'  banks  (central  creation) 
are,  ipso  facto,  solved.  In  particular,  we  readily  realize  that  a  bankers' 
bank  can,  still  less  than  a  member  bank,  afford,  or  have  an  interest,  to  be 
"loaned  up."  This  and  the  consequent  fact  that  central  banks  always 
kept  a  comfortable  margin  are  for  England  and  Germany  so  obvious  that 
it  would  be  waste  of  space  to  elaborate  them.  In  the  United  States,  both 
facts  are  blurred  by  their  very  recognition  by  legislation.1  The  tech- 
nique of  this  legislation,  however,  so  curiously  mixed  up  quantity  and 
safety  considerations  that  it  would  have  been  necessary  to  keep  a  surplus 
reserve,  even  if  the  legal  minimum  had  been  sufficient  to  meet  all  cases — 
which  a  rigid  figure  in  the  nature  of  things  can  never  be  as  long  as  it  is 
below  100  per  cent.  But  this  was,  owing  to  the  circumstances  just 
glanced  at,  both  a  difficult  and  a  costly  thing  to  do.  However,  it  is  worth 
while  noting  that,  quite  abnormal  situations  excepted,  the  New  York 
banks'  resources  were  never  exhausted  in  any  phase  of  the  cycle,  and  that 
they  almost  always  managed  to  keep  free  surplus  reserves,  though  only 
modest  ones — in  the  last  decade  before  the  war  they  averaged  for  the 
months  of  the  autumnal  drain  at  about  1.5  per  cent. 

This,  of  course,  quite  accords  with  our  view  of  the  actual  role  of  cen- 
tral banks  and  indirectly  teaches  something  about  it.  Their  position 

necessity  to  supervise  closely  the  member  banks  on  the  safety  and  success  of  which  the 
safety  and  success  of  those  bankers'  banks  themselves  largely  depended,  combined  with 
the  doubtful  quality  of  the  managements  of  many  of  the — especially  pigmy — member 
banks  to  create  a  situation,  the  natural  remedy  for  the  practical  difficulties  of  which  would 
have  been  mergers  plus  branch  banking.  The  merger  movement,  in  fact,  gathered  momen- 
tum from  1900,  when  the  amendment  to  the  National  Bank  Act  still  intensified  the  struggle 
between  larger  banks  for  member  banks'  balances  (see  chart  on  p.  358,  H.  P.  Willis  and 
J.  M.  Chapman,  The  Banking  Situation,  1984).  If  it  did  not  go  far  enough  to  strengthen 
the  credit  system  of  this  country,  that  was  due  to  hostile  legislation  and  to  an  attitude  of 
the  public  mind  which  would  not  have  stood  the  formation  of  units  big  enough  to  be  really 
safe  and  effective. 

1  Statutory  requirements  as  to  reserves  against  bank  notes  have  nothing  to  do  with  the 
above  argument. 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    651 

and  the  quantitative  weight  of  their  business  are  such  as  to  make  their 
mere  attitude  and  example  a  matter  of  moment  to  the  banking  com- 
munity and  to  the  business  world  in  general.  Rediscount  rate — or,  more 
generally,  the  conditions  at  which  they  stand  ready  to  finance  member 
banks — rationing,  suasion,  open -market  operations  undertaken  ad  hoc, 
need  be  mentioned  only,  to  call  up  roughly  appropriate  ideas  in  any  mind 
that  is  not  misled  by  theories  and  plans.  It  may,  however,  be  added  that 
rationing  was  not  an  emergency  measure  but  part  of  the  ordinary  routine 
of  bankers'  banks.  No  member  bank  was  allowed  to  rediscount  or 
borrow  all  it  may  have  wished  to,  but,  if  the  writer  may  judge  from  the 
practice  of  the  only  great  central  bank  which  he  thoroughly  knew  and 
from  such  indications  as  he  was  able  to  gather  about  the  practice  of  two 
others,  every  member  bank  was  closely  watched,  not  only  as  to  its  balance 
sheet,  but  also  as  to  its  personnel,  the  nature  of  the  transactions  it  entered 
into,  its  affiliations,  and  the  kind  and  quality  of  its  customers — whether, 
for  instance,  they  were  retailers  or  industrial  firms,  geographically  and 
economically  distributed  or  concentrated,  and  so  on.  Gossip  about  them 
and  their  leading  men  was  carefully  collected.  Thereupon,  it  was 
decided  what  its  "ration"  was  to  be.  This  was  then  varied  cyclically, 
besides  being  currently  revised  on  the  merits  of  the  individual  case.  By 
suasion,  as  distinct  from  (general)  attitude,  we  mean  attempts  to  influ- 
ence individual  member  banks  or  groups  of  member  banks.  It  covers  a 
wide  variety  of  things  and  would,  in  some  cases,  better  be  designated  as 
scowling  or  snarling.  It  ranges  from  threats — of  which  the  threat  to 
withhold  accommodation  is  only  one — and  admonitions  down  to  such 
measures  as  extending  or  withholding  invitations  to  official  dinners. 
Though  the  possibility  of  producing  effects  by  such  methods  greatly 
differs  from  country  to  country  and  from  time  to  time,  no  one  who  is  at 
all  familiar  with  the  working  of  credit  institutions  will  deny  that  it  is 
considerable.  This  alone  suffices  to  prove,  if  proof  be  wanted,  that 
bankers'  banks  cannot  harbor  any  systematic  tendency  toward  being 
•"loaned  up,"  and  that  the  argument  in  the  previous  section  applies  to 
them  with  added  force. 

It  would  be  quite  wrong  to  think  that  central  bank  action  is  merely 
adaptive  or  passive  in  any  sense  other  than  that  it  is  typically  action  in 
response  to  a  situation  not  itself  created  by  the  central  bank.  The  writer 
confesses  his  inability  to  see  the  point  at  issue  in  the  ancient  controversy 
whether  bank  rate  is  "declaratory"  or  "constitutive."  As  far  as  the 
bankers'  bank  also  transacts  member  bank  business,  it  cannot  help  being 
if  not  the,  at  least  a,  leader  in  an  imperfect  market.  As  far  as  it  does 
central  bank  business  only,  it  is  in  certain  cases  a  monopolist  and  in  all 
cases  in  a  position  in  which  any  of  its  moves  must  be  a  factor  in  the  sub- 
sequent situation.  Even  if  it  never  aimed  at  anything  else  but  at  fixing 
its  rate  in  such  a  way  as  merely  to  declare  or  register  what  the  financial 


652  BUSINESS  CYCLES 

situation  is,  we  should  have  to  recognize  that  the  man  who  has  it  in  his 
power  to  hoist  a  danger  signal  at  will  thereby  unavoidably  acquires  the 
power  to  create,  as  well  as  to  declare,  a  situation.  But  the  very  position 
in  which  a  bankers'  bank  finds  itself  makes  it  practically  impossible 
merely  to  register  the  state  of  things  such  as  it  is  at  the  moment,  still  less 
the  tendency  of  market  rates  alone. 

Central  banks'  rates  and  attitudes  owed  much  of  their  effectiveness 
to  the  fact  that  member  banks  and  the  public  looked  upon  them  as 
symptoms  of  the  situation  and,  to  a  considerable  extent,  reacted  accord- 
ingly. It  is  also  true  that  the  powers  of  central  banks  to  exert  mechanical 
effect  seemed — to  economists  as  well  as  to  business  men — to  be  greater 
than  they  really  were.  Yet  the  means  at  the  command  of  central  banks 
to  exert  such  effects,  sometimes  acting  upon  sensitive  margins,  were 
obviously  adequate  to  managing  any  ordinary  situation  and  also  to 
securing  for  them  considerable  freedom  of  action,  though  much  more  of 
course  in  the  case  of  single  central  banks  than  in  the  case  of  competitive 
bankers'  bank  systems.  Criticism,  which  has  become  traditional,  draws 
a  different  picture — mainly,  it  seems,  for  three  reasons. 

First,  most  critics  concentrate  attention  on  exceptional  situations. 
It  is,  no  doubt,  relevant  to  the  analysis  of  the  functioning  of  any  central 
bank  or  central  banking  system,  to  see  how  it  lived  in  rough  weather. 
But  to  make  this  the  only  criterion  implies  overlooking  all  the  problems 
to  be  solved  in  the  normal  run  of  things,  which  it  is  easy  to  overlook 
precisely  because  they  were  solved  successfully.  Second,  it  has  been 
urged  that  central  banks  reacted  mechanically  to  the  behavior  of  certain 
indices  relevant  to  their  own  situation  and  perhaps  that  of  the  money 
market  and  of  the  currency,  without  any  regard  to  the  economic  organism 
of  their  nations  as  a  whole  and  without  any  conception  of  the  therapeutic 
influence  they  might  have  exerted.  But  if  we  consider  what  it  was  that 
the  Bank  of  England  was  supposed  to  react  to — reserve  proportion,  gold 
movements,  and  so  on — it  becomes  evident  that  in  the  vast  majority  of 
cases  (and  in  exceptional  situations  neither  the  Bank  of  England  nor  the 
Reichsbank  acted  upon  such  indices  alone),  diagnosis  of  the  cyclical 
situation  of  a  country  would  have  coincided  with  the  inference  that  any 
outside  observer  could  have  drawn  from  those  indices,  and  that  action 
upon  that  inference  must,  hence,  have  had  some  "stabilizing"  effect  on 
volume  of  transactions,  prices,  and  so  on.  Finally,  however,  most 
critics  speak  from  the  background  of  a  banking  theory  of  the  business 
cycle  and  imply  that  central  banks  should  be  able  to  iron  it  out  entirely. 
Such  a  misconception  of  course  negatives  all  possibility  of  realistic 
analysis  and  substitutes  a  spurious  problem  for  the  real  one. 

It  is  at  the  same  time  obvious  from  the  nature  of  the  means  by  which 
central  banks  may  influence  economic  processes  that  there  are  definite, 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    653 

though  elastic,  limits  to  their  effectiveness,  and  that  those  means  go 
much  further  if  the  intention  is  to  put  on  brakes  than  if  the  intention  is  to 
stimulate.  That  this  was  so  as  a  matter  of  historical  fact  cannot  well  be 
doubted.  But  it  is,  again,  a  distinct  question  whether  this  was  so  merely 
by  virtue  of  the  ideas  about  sound  currency  that  framed  the  legal  powers 
and  duties  of  central  banks,  their  traditions,  and  practices,  or  because  of 
more  fundamental  reasons.  As  we  have  seen  in  the  case  of  member 
banks,  that  legal  or  customary  restrictions,  though  they  may  be  prac- 
tically unavoidable  to  curb  recklessness,  do  not  create  the  necessity  for 
that  attitude  of  restraint  that  is  inherent  in  the  conditions  of  banking, 
so  we  have  no  difficulty  in  recognizing  in  the  case  of  central  banks,  that 
an  analogous  attitude  would  impose  itself  on  them  by  virtue  of  the 
logic  of  their  situation,  even  if  no  legislation  or  tradition  enforced  it. 
The  main  obstacle  that  prevents  the  adherents  of  the  most  varied 
"schools"  to  see  this  is  the  argument  that  a  central  bank,  being  the 
"ultimate  creator  of  credit,"  is  exempt  from  the  limitations  banking 
practice  imposes  on  member  banks,  and  hence,  in  the  absence  of  those 
legal  restrictions  and  traditions,  would  enjoy  almost  unlimited  freedom 
in  acting  on  business  situations.  Quite  apart  from  the  fact  that  govern- 
ment could  never  afford  to  allow  it  to  fail,  it  need  not  bother  about 
"quality"  and  "purpose"  of  credit  at  all,  since  it  would  have  it  in  its 
power  to  improve  any  "quality"  and  to  justify — in  the  business  sense — 
any  "purpose  "  to  any  desired  extent  by  further  creation  of  balances.  We 
will  discuss  this  with  reference  to  the  sequence  of  cyclical  situations  in  an 
isolated  domain,  although  our  material  does  not  permit  the  elimination 
of  the  complications  that  arise  from  international  trade,  gold  movements, 
foreign  exchanges.  At  every  step,  problems  of  actual  and  of  possible 
behavior  will  be  kept  distinct.  The  factor  of  redcemability  of  balances 
in  a  legal  tender  that  also  serves  as  cash  enters,  of  course,  into  the  former 
only. 

As  pointed  out  above,  the  central  bank  may,  owing  to  its  conspicuous 
position,  exert  some  influence  on  the  temper  of  the  business  community, 
which  at  certain  critical  junctures  may  be  considerable — especially  in 
preventing  or  stopping  panics — but  ordinarily  cannot  go  further  than 
hortatory  efforts  by  leading  officers  of  state,  which  are  immediately  put 
into  their  place  by  the  recognition  of  the  truth  that  cyclical  phases  are 
more  than  mere  matters  of  psychology.  Excepting  this  and  the  influence 
exerted  by  a  central  bank's  member  bank  business  with  firms  and  house- 
holds, it  is  clear  from  the  outset,  however,  that  its  action  in  the  central 
market  affects  the  money  and  the  open  markets  primarily  through  mem- 
ber banks  reserves,1  and  hence  the  business  processes  beyond  those  mar- 

1  It  is  not  required  that  the  regulation  of  member  bank  reserves  be  a  definite,  clearly 
perceived,  and  theoretically  understood  aim  of  the  managers  of  the  central  bank.  Such 


654  BUSINESS  CYCLES 

kets  only  insofar  as  variation  in  these  reserves — supplemented  by  atti- 
tude and  suasion — can  influence  them.  But  this  means,  first,  that,  even 
if  central  banks  were  so  completely  masters  of  member  bank  systems  that 
their  slightest  move  would  be  immediately  translated  into  an  increase  or 
decrease  of  members'  willingness  to  lend,  no  greater  effect  on  industry 
and  trade  could  be  expected  than  we  have  seen  reason  to  expect  from 
variations  in  this  willingness:  one  only  of  the  data  on  which  firms  act 
being  affected  thereby,  definite  results  could  be  predicted  only  under 
ceteris  paribus  assumptions,  perfectly  inadmissible  within  our  process. 
Second,  even  for  member  banks  themselves  increased  and  decreased 
power  to  lend  is  only  one  of  many  factors  that  determine  their  willingness 
to  lend,  and  the  decisive  one  in  some  situations  only  and  not  in  all. 
Member  banks  are  distinct  centers  of  economic  decision.  No  policy  of 
the  central  bank,  short  of  a  declaration  to  sanction  and  make  good  the 
consequences  of  any  action  whatsoever,  can  alter  all  the  data  on  which 
such  decisions  rest  and,  in  particular,  the  logic  of  the  banking  business 
described  in  the  preceding  section.  It  would,  hence,  not  even  be  correct 
to  say  that  the  central  bank  determines  the  "supply  schedules" — if  this 
expression  were  permissible — of  member  banks,  although  it  influences 
them  by  its  greater  or  lesser  readiness  to  shoulder  the  burden  of  tensions 
that  may  arise.  The  analogy  between  the  relation  of  a  member  bank 
to  its  industrial  customers  and  the  relation  of  a  bankers'  bank  to  mem- 
ber banks  is,  therefore,  not,  as  one  might  think,  destroyed  if  the  bankers' 
bank  is  the  ultimate  creator  of  credit  in  the  sense  in  which  a  single  central 
bank  may  be  said  to  be:  inferences  from  the  associations  which  that 
phrase  is  likely  to  call  forth  would  be  completely  misleading.  Nor  do 
open-market  operations  alter  the  argument.  It  has  been  often  observed, 
and  is  indeed  a  patent  fact,  that  member  banks  occasionally  thwart  the 
action  of  the  central  bank — a  qualification  will,  however,  be  added  to  this 
in  Chap.  XIV,  Sec.  F,  V — and  that  the  latter  sometimes  proves  to  be  a 
"dud."  This  is  perfectly  understandable  from  our  standpoint  and  is  no 
deviation  from,  but  part  of,  the  blueprint  of  the  credit  machine.  The 
fact  that  central  bank  action  sometimes  does  and  sometimes  does  not  take 
effect  is  a  problem — and  an  awkward  one — only  for  believers  in  the 
mechanical  action  on  business  of  mere  volume  of  credit  facilities.  What 
central  bank  policy  can  be  said  to  determine  directly — pari  passu  with 
members'  reserves — is  the  central-market  rate.  It  also  very  powerfully 
acts  on  open-market  rates.  But  there  is  a  long  way  from  action  in  the 
central  market  to  action  on  the  money  market  in  our  sense,  and  a  still 

regulation  will  in  many  cases  be  implied  in  their  ordinary  business  behavior.  The  Bank 
of  England  exerted  regulating  influence  of  this  kind  on  country  and  on  London  banks,  at 
least  from  the  beginning  of  the  nineteenth  century,  without  developing  any  theory  or 
principle  of  it  for  a  long  time.  Open-market  operations  are  about  as  old. 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    655 

longer  one  from  this  to  action  on  business  activity  and  prices.  Even 
legally  unlimited  power  to  create  balances  does  not  imply  actual  power  to 
create  them,  still  less  the  power  to  make  them  active. 

Since  central  banks  mainly  act  on  business  through  member  banks, 
it  is  unnecessary  to  show  that  the  former,  no  more  than  the  latter,  "cause" 
the  system's  prosperity  excursions,  though  they  could  conceivably  prevent 
them.  In  strict  theory,  particularly  if  the  preceding  neighborhoods  come 
sufficiently  near  to  being  equilibrium  states,  central  bank  action  at  such 
times  may  with  equal  justification  be  said  to  be  absent  or  neutral  or  con- 
fined to  those  small  adjustments  of  the  steering  wheel  that  are  necessary 
even  on  the  straightest  course  in  the  calmest  of  seas.  In  the  absence  of 
effects  of  external  factors  or  of  phases  of  underlying  cycles,  there  is,  during 
this  situation,  as  much  encouragement  to  member  banks  in  the  central 
bank's  attitude  and  rate  as  there  is  encouragement  to  entrepreneurs  in 
member  banks'  attitudes  and  rates,  and  no  more — there  is  hence  no 
Wicksellian  effect  from  any  discrepancy  between  money  and  real  rate  of 
interest.  The  presence  of  external  factors — capital  and  gold  movements, 
in  particular — may  alter  all  this,  and  it  is  in  connection  with  these  that 
the  true  role  of  central  bank  initiative  reveals  itself.  But  this  we  dis- 
regard for  the  moment.  As  prosperity  sets  in,  the  central  bank  typically 
watches  the  expansion  of  business  volume  and  of  banking  activity, 
slowly  raising  its  rate  in  response.  Normally  there  is  at  this  stage  no 
motive  to  exert  restrictive  initiative,  nor  would  it  be  easy  to  do  so,  both 
because  of  the  pressure  of  public  opinion  and  because  of  the  fact  that  the 
money  market  is,  as  we  have  seen,  not  only  sloppy  in  the  sense  in  which 
the  market  of  any  commodity  may  be,  but  in  the  special  sense  that  its 
funds  are  normally  underutilized  in  and  immediately  after  a  neighborhood 
of  equilibrium,  so  that  any  practically  feasible  attempt  at  restriction 
would  be  thwarted  by  members,1  and  nothing  but  the  most  drastic  open- 
market  operations  could  avail — borrowing  unlimited  amounts  at  10  per 
cent,  for  instance. 

As  prosperity  wears  on,  however,  and  member  banks  liquidate  part  of 
their  secondary  reserve — calling  demand  loans  to  the  stock  exchange  and 
to  bill  brokers — the  rising  bank  rate  tends  to  become  effective,  both  in 
the  sense  that  it  applies  to  a  larger  volume  of  business  which  the  tighten- 
ing of  the  open  market  drives  towards  the  central  bank,  and  in  the  sense 
that  it  exerts  restraining  or  punitive  influence.  In  London,  for  instance, 
more  and  more  of  the  bill  brokers'  material  had,  at  that  stage,  to  be 
offered  at  the  Bank  for  rediscounting.  This  theoretically  goes  on,  and  in 

1  In  practice  member  banks  could  also  thwart  a  policy  of  stringency  which  the  central 
bank  may  be  trying  to  impose,  by  shifting  balances  among  themselves  and  drawing  funds 
from  abroad.  Issues  of  stocks  and  bonds,  however,  primarily  act  in  the  direction  in  which 
the  central  bank  wishes  to  steer.  A  policy  of  restriction  may,  in  fact,  be  intended  to  enforce 


656  BUSINESS  CYCLES 

practice  mostly  did  go  on,  until  the  situation  reversed  itself  and  the 
economic  process  settled  into  the  perfectly  normal  and  anything  but 
catastrophic  processes  of  recession.  There  was  no  spectacular  manage- 
ment about  this,  such  as  ardent  stabilizers  wish  to  see,  but  it  was  mone- 
tary management  involving  a  lot  of  general  guidance  of  things,  and 
central  banks,  such  as  the  Bank  of  England  and  the  Reichsbank,  can  no 
more  be  said  to  have  mechanically  followed  events  than  a  horseman,  who 
refrains  from  spectacular  whippings  and  spurrings  and  acts  on  his  mount 
primarily  by  small  adjustments  of  his  seat,  can  on  that  account  be  said  to 
"follow"  his  horse.  Nor  were  there  at  those  junctures  any  sudden  jerks 
caused  by  central  banks'  running  up  against  technical  limits  and  causal  to 
the  occurrence  of  the  upper  turning  point — which  could  be  compared,  if 
we  wish  to  keep  to  our  analogy,  to  brutal  pulls  at  the  curb,  productive 
of  discomfiture  to  horse  and  rider. 

This  is  what,  in  the  absence  of  external  disturbances,  central  banks 
actually  did,  although  the  picture  would  in  its  details  have  to  differ  for 
different  periods  and  countries,  in  particular  according  to  whether  central 
banks  directly  rediscounted  for  member  banks  or  not.  It  should  be 
observed,  first,  that  in  important  points,  though  not  in  all,  our  sketch 
supports  the  classic  theory  of  central  banking.  But  it  could  almost 
equally  well  be  expressed  in  terms  of  more  up-to-date  theories  of  monetary 
management.  The  objection  to  both  is  that  they  remain  on  the  surface. 
But  as  far  as  they  go,  the  real  difference  between  them  does  not  justify 
the  emphasis  put  upon  it.  Second,  it  should  be  noticed  that  central 
bank  behavior  as  outlined  also  conforms  to  the  rules  the  profit  motive 
would  dictate.  Only,  they  would  have  to  be  defined  not  by  the  principle 
of  maximization  of  instantaneous  profits  but  with  reference,  on  the  one 
hand,  to  the  general  logic  of  banking  and,  on  the  other  hand,  to  the  fact 
that  a  concern,  situated  as  the  central  bank  is,  can  never  fail  to  link  its 
interest  with  the  state  of  the  whole  economic  organism.  Third,  the 
formal  analogy  may  be  mentioned  which  exists  between  the  cyclical 
and  the  seasonal  situations  that  central  banks  have  to  face.  When,  in 
particular,  prosperity  is  under  way  they  face  a  situation  technically 
similar  to  that  which  first  attracted  the  attention  of  economists  under  the 
heading  of  Autumnal  Drain.1  This  and  similar  seasonal  variations  in 
the  volume  of  borrowing  and  in  the  flow  of  cash  (income  tax  payments, 
Christmas  business,  and  so  on)  bankers'  banks  learned  how  to  handle, 
although  the  fact  that  crises  broke  more  often  at  times  of  seasonal  strain 
than  at  other  times  persisted  throughout  the  period.  In  England  there 
was  no  significant  association  between  bank  rate  and  either  London 
clearinghouse  figures  or  circulation  (notes  plus  coin)  within  each  year, 

1  Cf.  W.  S.  Jevons,  On  the  Frequent  Autumnal  Pressure  in  the  Money  Market,  Journal 
of  the  Royal  Statistical  Society,  1866. 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    657 

although  there  was  between  seasonal  gold  movements  into  and  out  of  the 
Bank  of  England,  bank  rate  and  the  price  of  gold.  In  Germany  the 
handling  of  seasonal  strains  was  facilitated  by  special  legislation  (increase 
of  tax-free  amount  of  bank  notes  issued  in  order  to  cope  with  quarterly 
requirements;  no  provision,  however,  was  made  for  the  autumnal  drain). 
In  the  United  States  the  experience  of  1907  exerted  some  pedagogical 
influence  that  relieved  seasonal  tensions.  But  more  important  than  the 
analogy  is  the  difference  between  seasonal  and  cyclical  problems — the 
former  presenting  nothing  but  the  clear-cut  task  of  minimizing  incon- 
venience; the  latter,  the  much  more  formidable  one  of  dealing  with  a 
fundamental  economic  process. 

What  central  banks  could  do  if  they  were  freed  from  technical  fetters 
and,  in  particular,  from  the  pressure  of  currency  movements  and  of 
foreign  exchanges  is  another  matter.  It  has  been  stated  above  that  they 
could,  in  the  limiting  case,  not  only  prevent  all  member  creation  but, 
by  taking  up  all  available  funds,  prevent  prosperities  themselves.  It  is 
no  less  clear  that  in  the  atmosphere  of  high  prosperity,  with  excesses 
everywhere  in  full  swing,  expanding  member  bank  reserves  would  induce 
increased  lending  and  take  effect  in  the  business  world.  Hence,  by 
unreservedly  standing  for  indefinite  inflation  bankers'  banks  no  doubt 
could  defer  occurrence  of  the  upper  turning  point  indefinitely,  i.e.,  until 
the  monetary  system  breaks  down.  Unless  this  be  an  end  in  itself, 
there  seems,  however,  no  point  in  trying  to  fight  recession  when  it  sets  in 
if  we  remember  what  it  normally  means.  We  have  here  the  final  reason 
why  any  really  or  hypothetically  unlimited  power  of  central  banks  to 
create  credit  makes,  practically  and  analytically,  less  difference  than  one 
would  suppose. 

When,  in  the  course  of  recession,  business  normalizes  itself  and  liquid- 
ity increases  all  round,  central  banks'  control  over  member  banks' 
reserves  and  the  open  market  and  their  indirect  influence  on  business 
become  progressively  weaker.  The  horse  is  no  longer  running  up  to  his 
bit.  But  in  the  absence  of  international  complications  this  does  not 
normally  create  any  problem  necessitating  interference  by  restrictive 
open-market  operations.  On  the  contrary,  central  banks  are  in  this 
phase  both  able  and  willing  to  give  rein  and  even  to  urge  on  by  a  low 
rediscount  rate,  although  no  initiative  on  their  part  is  as  a  rule  called  for. 
This  statement,  which  implies  that  depression  (not  crises)  should  set  in 
while  conditions  of  easy  money  prevail,  hence  not  because  of  monetary 
stringency,  can  sound  paradoxical  only  to  adherents  of  monetary  theories 
of  the  cycle — some  of  whom  will,  it  is  hoped,  be  converted  by  the  obvious 
facts  of  1937  to  1938 — and  amounts  to  saying  that  central  banks  can  do 
little  beyond  what  their  admonitions  may  effect  in  order  to  prevent 
depressions.  What  can  be  done  at  all  from  the  banking  angle,  member 


658  BUSINESS  CYCLES 

banks  should  in  that  situation  be  able  to  do  themselves.  This  is  almost, 
but  not  quite,  independent  of  the  legal  fetters  imposed  on  central  creation, 
for  if  there  were  absolutely  no  limits  to  the  latter  and  if  the  central  bank 
made  itself  unconditionally  responsible  for  whatever  transactions  people 
were  prepared  to  enter  into  and  undertook  to  finance  all  ensuing  deficits 
ever  after — otherwise  a  crisis  would  immediately  follow  upon  the  with- 
drawal of  the  guarantee — this  would  no  doubt  produce  effects. 

For  the  same  reason  and  with  the  same  qualification,  there  is  not 
much  room  for  central  bank  initiative  in  the  course  of  depression.  If 
facilities  were  forced  upon  them  by  open-market  buying,  member  banks 
would  thwart  the  intention,  first,  by  using  their  access  of  funds  in  order  to 
repay  their  debts,  and  then  by  accumulating  secondary  or  simply  idle 
reserves.  Moreover,  it  follows  from  our  diagnosis  of  the  nature  of 
depression  that,  as  far  as  they  could  be  prevailed  upon  to  accept  against 
their  judgment  such  business  as  they  may  be  able  to  attract  under  the 
circumstances,  this  would  be  a  source  of  additional  difficulties  in  the 
future.  But  there  is  another  way  in  which  the  latter  effect  may  follow 
from  an  attempt  of  bankers'  banks  to  enforce  expansion  of  the 
volume  of  balances  during  depression.  It  is  obvious  how  difficult, 
if  not  "politically  impossible,"  it  must  be  to  retrace  such  steps.  The 
opportune  moment  for  doing  so  will,  in  the  eyes  of  the  member  banks  and 
of  the  public,  never  come.  The  delicate  processes  of  incipient  revival 
could,  in  fact,  be  easily  stalled  by  anything  looking  like  restriction,  and 
public  resistance  is  not  likely  to  decrease  later,  particularly  if  member 
banks  have  accumulated  not  idle  but  secondary  reserves.  Thus  the 
system  might  reach  the  neighborhood  of  equilibrium  in  a  state  of  abnor- 
mal liquidity,  and  in  the  subsequent  prosperity  the  central  banks'  action 
in  the  previous  depression,  or  the  member  banks'  surplus  reserves  created 
thereby,  will  indeed  produce  effects,  viz.,  speculative  excesses,  reckless 
banking,  an  overgrown  Secondary  Wave,  and  later  on,  breakdowns.  It  is 
this  vicious  effectiveness  of  therapeutic  efforts  of  this  type  rather  than 
the  mere  futility  of  trying  during  depression  to  drown  pessimism  in  a 
flood  of  credit  that  should  be  stressed  in  arguments  about  it.  We  may 
note  in  passing  that  permanent  expansion  of  the  circulating  medium 
may  in  this  way  ensue,  and  that  fluctuations  of  prices  may  thenceforth 
be  permanently  from  a  level  higher  than  would  otherwise  obtain. 
Gold  discoveries  act  in  the  same  way  rather  than  by  directly  creating 
prosperities. 

It  is  recovery  which,  of  all  the  phases  of  an  ordinary  cycle,  presents  the 
most  difficult  and  most  important  practical  problems  for  bankers'  banks. 
The  situation  may  so  clear  up  in  the  last  stages  of  depression  that  the 
responsibility  can  be  taken  for  a  lead  that  may  help  to  bring  about  the 
lower  turning  point  more  quickly  without  producing  undesired  effects 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    659 

as  well  as  the  desired  one.  But  for  reasons  we  need  not  repeat,  such  a 
lead  is  likely  to  be  followed  much  more  promptly  by  those  sectors  of  the 
community  which  are  likely  to  overdo  things  than  by  the  ordinary  run  of 
"legitimate"  business.  And  this  difficulty  grows  in  importance  in  the 
later  stages  of  recovery,  when  everyone  realizes  the  upward  tendency 
and  is  disposed  to  press  forward  in  any  case.  The  necessity  for  steadying 
advance  thus  soon  becomes  more  evident  than  that  for  propelling  it, 
and  we  shall  not  be  surprised  to  find  that  central  banks  were,  in  revivals, 
as  a  rule  more  concerned  with  controlling  general  liquidity  than  with 
trying  to  add  to  it,  and  to  take  an  attitude  which  the  public  very  naturally 
disapproved.  Under  the  general  conditions  of  revival,  the  money  and 
open  markets  were  most  likely  to  get  out  of  hand,  and  it  was,  hence, 
primarily  the  experience  with  the  excesses  and  vicissitudes  that  fre- 
quently occurred  in  revivals — Juglar  revivals,  in  particular — which 
eventually  led  to  an  almost  general  demand  for  measures  calculated  to 
strengthen  central  bank  control.  It  follows  that  central  banks  could  have 
done  more  than  they  did  to  stabilize  things  in  revivals.  But  the  main 
obstacle  to  this  was  public  opinion,  although  until  the  end  of  the  seventies 
neither  the  task  nor  the  technique  was  perfectly  understood.  They  also 
could,  if  freed  from  technical  fetters  and  from  any  other  consideration 
except  how  to  produce  booms,  have  accelerated  the  processes  of  revival 
at  the  expense  of  producing  corresponding  slumps. 

Nothing  has  been  said  about  how  central  banks  did  behave  or,  to 
retain  our  double-track  argument  to  the  end,  could  have  behaved  in 
crises.  In  order  to  bring  out  the  fundamental  question,  the  above  analy- 
sis has  been  conducted  within  the  walls  of  assumptions  that  not  only 
excluded  the  problems  incident  to  international  relations — gold  move- 
ments in  particular — and  to  domestic  disturbances  of  a  non-cyclical 
nature  but,  as  much  as  possible  also,  the  abnormal  features  of  the  cyclical 
process  itself.  Within  these  restrictions  we  have  seen  that,  by  guiding 
and  managing  all  the  time,  central  banks  allowed  the  cyclical  process  to 
take  its  course,  taxing  expansion  in  the  later  stages  of  prosperity,  but 
did  not  by  their  action  create  any  of  the  cyclical  phases;  that  such  guid- 
ance had  also  another  rationale  than  mere  regard  to  golden  fetters  would 
imply;  that  it  rarely  if  ever  required  that  bank  rate  should  precede  in  the 
time  sequence  of  the  factors  of  business  situations  and  that  its  failure  to 
do  so  did  not  cause  or  aggravate  slumps;  that  by  its  nature  and  not  only 
by  statute  or  tradition,  regulation  primarily  meant  restrictive  regulation; 
but  that  this  regulation  was  never  mechanical  or  uniquely  determined  by 
obedience  to  a  few  indices.  And  it  would  be  easy  to  understand  on  these 
considerations  alone  that  bank  rate  was,  in  both  England  and  Germany, 
regularly  above  the  rates  of  the  open  market.  But  it  might  be  urged 
with  some  justice  that  the  behavior  of  such  a  regulating  agency  cannot  be 


660  BUSINESS  CYCLES 

described  in  terms  that  exclude  precisely  those  irregularities  which  accord- 
ing to  some  economists  primarily  call  for  regulation. l  Within  an  isolated 
domain,  the  relation  of  central  banks  to  what  we  call  reckless  banking, 
speculative  excesses,  fraudulent  or  irresponsible  business  activity — 
especially,  finance — is  the  one  of  the  two  most  important  points,  and  the 
treatment  of  crises  or  panics  is  the  other. 

Dealing  with  the  former  would,  in  order  to  be  effective,  require  a 
policing  power,  which  to  this  day  has  always  been  quite>  beyond  central 
banks.  This  inability  of  capitalism  to  police  itself  is  as  striking  as  its 
inability  to  protect  itself — it  always  requires  both  a  policeman  and  a 
protector  of  nonbourgeois  complexion,  who  regulate,  shield,  and  exploit 
it.  This  is  as  true  of  the  times  of  Queen  Elizabeth  as  it  is  today.  But 
it  is  largely  this  inability  that  produces  crises  as  distinguished  from  mere 
depressions.  For  the  prewar  period,  at  least,  historical  evidence  is 
conclusive  on  this  point.  As  soon  as  we  accept  this  evidence  we  had 
better  stop  talking  about  the  causation  of  crises  in  terms  of  bank  rate  or 
inadequate  accommodation  imposing  restrictions  that  were  relaxed  only 
"when  the  damage  had  been  done"  and  that  turned  into  catastrophes 
what,  with  another  central  bank  policy,  would  have  been  perfectly 
normal  situations.  Crises  would  have  occurred  if  no  central  or  member 
bank  had  ever  called  a  single  loan,  and  must  be  understood  in  the  light 
of  the  fact  that  in  any  economic  and  social  system  which  is  unable  to 
prevent  irresponsibility  and  misconduct  correction  by  consequences  is 
the  only  method  to  prevent  indefinite  aberration.  At  least  in  historical 
retrospect  and  on  the  understanding  that  from  this  recognition  of  patent 
facts  alone  no  practical  conclusions  can  be  drawn  as  to  possibilities  and 
desirabilities  of  later  times,  this  should  be  admitted,  however  much  we 
may  understand,  and  sympathize  with,  all  the  strata  which  that  method 
victimized.  Then  both  the  initiative  more  or  less  regularly  taken  by 
central  banks  in  such  situations  and  its  limitations  will  appear  in  their 
proper  light.  Their  position  was  such  that  the  social  meaning  of  crises 
translated  itself  for  them  into  obvious  business  considerations.  Owing 
to  their  inadequate  power  and  will  to  police  the  world  of  banking  and 
finance,  their  action  came,  indeed,  unavoidably  after  the  event — this  is 
the  only  sense  in  which  it  can  be  held  that  it  came  "too  late,"  and  this 
sense  has  nothing  to  do  with  bank  rate.  Coming  when  it  did  it  had  to  be 
remedial  and  not  punitive.  Success  went  in  some  cases,  which  we  met 
in  our  historical  sketch,  to  the  length  of  preventing  panics  and  that 

1  Although  the  present  writer  does  not  go  so  far  as  that,  but  admits  that  even  for  the 
most  normal  course  of  things  there  is  a  special  case  for  monetary  as  distinct  from  other 
regulation,  it  should  be  pointed  out  that  the  difference  between  the  views  presented  above 
and  some  that  are  at  present  more  widely  held  is  partly  due  to  the  fact  that  we  deal  here 
with  one  particular  aspect.  Many  views  on  monetary  management,  for  instance,  that 
seem  to  differ  fundamentally  from  ours  primarily  refer  to  the  international  gold  situation. 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    661 

restriction  of  credit  which  will  ensue  in  panics.  In  most  other  cases,  it  is 
easy  to  see  that  consequences  were  substantially  mitigated  and  durations 
of  spirals  shortened,  as  compared  with  what  we  may  conceive  they  would 
otherwise  have  been,  by  the  initiative  action  taken  by  central  banks. 

The  general  logic  of  their  position  shows  in  such  situations,  as  it  were 
through  a  magnifying  glass.  They  obviously  could  not  simply  stand  for 
saving  everything  and  everybody,  because  this  would  have  meant  con- 
doning error  and  misconduct  to  the  point  of  seriously  impairing  the 
efficiency  of  the  system  with  which  they  had  to  work.  Saving  the 
Gurneys,  for  example,  would  have  meant  saving  ironworks  and  shipping 
ventures  which  could  not  have  lived  without  permanent  subsidies.  But 
within  the  limits  set  by  this  consideration  the  helping  hand  was  freely 
extended.  In  the  English  case  in  particular,  the  suspensions  of  Peel's 
Act  must  be  interpreted  not,  as  contemporaneous  and  later  criticism 
often  interpreted  them,  viz.,  as  so  many  breakdowns  of  the  engine,  but  as 
parts  of  it  that  were  put  into  operation  in  order  to  show  in  moments  of 
panic  the  "ultimate  creator  of  credit"  in  all  the  glory  of  unlimited  power. 
That  engine  was  not  foolproof,  to  be  sure,  and  the  way  in  which  it  was 
operated  at  any  given  point  of  time  and  in  which  it  was  made  to  deal 
with  situations  that  were  essentially  historical  individuals  may  on  a 
different  level  of  analysis  give  plenty  of  scope  for  criticism,  quite  inde- 
pendent of  any  theories  about  the  magic  possibilities  of  bank  rate.  But 
this  does  not  affect  the  question  of  principle,  with  which  we  are  concerned. 

Thus,  while  we  need  not  stay  to  discuss  the  obvious  arguments  that 
may  be  adduced  against,  or  the  equally  obvious  case  that  may  be  made 
out  for,  that  institutional  arrangement,  or  the  question  how  its  sources 
of  waste  compare  with  those  of  the  Gosplan,  it  is  necessary  to  emphasize 
that  within  its  framework  central  banks  could  hardly  have  done  much  more 
than,  for  example,  the  Bank  of  England  or  the  Rcichsbank  actually  did. 
Control  over  business  beyond  those  limits  cannot  be  usefully  discussed  in 
terms  of  banking  policy  but  only  in  terms  of  a  much  more  thoroughgoing 
type  of  management  of  the  underlying  industrial  and  commercial  proc- 
esses, which  would  call  for  agencies  of  different  construction.  In  this 
sense  bank  reform  is  a  technical  and,  moreover,  an  intracapitalist  matter, 
which  cannot  be  dealt  with  in  an  extracapitalist  spirit  and  which  has 
little  to  do  with  the  fundamental  issues  of  today. 

Chart  XXXV  is  presented  in  order  to  illustrate  the  results  of  this 
analysis.  The  behavior  of  Notes  in  the  Hands  of  the  Public,  of  course, 
reflects,  besides  the  cyclical  process,  the  secular  or  "structural"  change 
that  occurred  in  their  role  in  the  monetary  system.  A  similar  remark 
applies  to  Private  Deposits,  which  otherwise,  together  with  the  ratio  of 
Banking  Department  Reserve  to  Deposits  and  Bank  Post  Bills  and  with 
Other  Securities,  conform  to  expectation  from  our  argument  to  an  extent 


662 


BUSINESS  CYCLES 


which  is  remarkable,  considering  how  many  external  factors  there  were  to 
produce  deviations.  If  the  reader  looks  at  the  chart  in  the  light  of  our 
historical  report,  he  will  easily  satisfy  himself  that  our  process  substan- 


1850  I860  1870  1880  1890  1900  1910 

CHART  XXXV.— Bank  of  England  and  Allied  Series  (see  Appendix,  p.  1063). 

tially  suffices  to  account  for  the  behavior  of  those  series.  London  Total 
Clearings  and  the  ratio  between  them  and  Total  Deposits  (of  joint  stock 
and  private  banks  as  reported  by  the  Economist)  facilitate  interpretation 
since  the  middle  nineties.  The  ratio  between  London  Bankers'  Balances 
and  Reserve  obviously  contains  a  special  trend. 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    663 

How  far  our  instance  of  a  competitive  system  of  bankers'  banks,  the 
national  banks  of  New  York  City,  qualifies  for  the  role  may  in  part  be 
inferred  from  Chart  XXXVI.  It  should  be  added  that  the  growth,  as 
evidenced  by  the  increase  in  capital  and  surplus,  of  New  York  national 


1875  1880  1885  1890  1895  1900  1905 

CHART  XXXVL— United  States  (see  Appendix,  p.  1064). 


1910 


1914 


banks  after  the  reform  of  1900 — much  stronger  than  the  increase  in 
capital  and  surplus  of  outside  national  banks — was  precisely  due  to  their 
success  in  attracting  deposits  of  national  and  state  banks  outside  New 
York  and,  hence,  to  bankers*  bank  business.  Study  of  the  chart  reveals 
the  fact,  pointed  out  before,  that  both  net  deposits  and  loans  of  New  York 
banks  moved  in  the  short  run  inversely  to  the  deposits  and  loans  of  out- 


664  BUSINESS  CYCLES 

side  banks.1  This  implies  inverse  association,  in  the  short  run,  of  New 
York  loans  and  commercial  paper  rate.  As  has  also  been  mentioned 
already,  there  is  less  difference  in  behavior  in  the  case  of  investments, 
though  the  strongly  inverse  short-run  association  between  outside  loans 
and  outside  investments  is  not  present  in  the  New  York  figures,  and  in 
the  case  of  reserve  money  held.  The  distribution  of  money,  "  lawful "  and 
other,  between  New  York  and  outside  banks  does  not  seem  to  have 
played  any  considerable  role  in  the  mechanism  of  short  cycles  in  this 
country.  All  this  substantially  holds  also  for  Juglar  fluctuations,  but 
in  the  largest  contours  the  differences  are  naturally  much  less  in  evidence. 

We  conclude  that  New  York  banks  in  fact  acted  as  bankers'  banks. 
Of  course  it  must  be  borne  in  mind  that  bankers'  bank  business  was  only 
part  of  their  total  activities.  But  it  is  nevertheless  clear  that  the  varia- 
tions in  their  loans  and  investments  were  a  function  of  the  flow  of  money 
— both  of  cash  and  of  deposits — from  and  to  their  "member  banks," 
which  cyclically  swelled  and  depleted  their  deposits.  Interpretation  is, 
hence,  exactly  opposite  to  what  it  was  in  the  case  of  outside  banks,  in 
the  one  case  deposits,  in  the  other  case  loans  being  the  primary  phe- 
nomenon.2 Net  deposits  reflecting  the  variations  in  the  difference 
between  the  amounts  due  to  and  due  from  other  banks  were,  in  New 
York,  persistently  and  considerably  larger  than  loans  and  fluctuated  more 
strongly  than  in  outside  banks. 

The  competitive  character  of  this  bankers'  bank  system,  of  course, 
deprived  its  component  units  of  much,  if  not  all,  the  freedom  that  a  single 
central  bank  enjoys.  Much  less  power  for  initiative  action  was  left  to 
them,  and  they  were  often  compelled  to  sail  very  close  to  the  wind.  As  in 
every  fall  deposits  were  withdrawn  and  cash  streamed  out  into  the  West 
and  South,  so  in  every  prosperity  an  analogous  phenomenon  asserted 
itself  and  forced  them  to  liquidate  their  temporary  investments  and  even 
to  take  gold  from  Europe  against  bills  drawn  in  payment  for  exports  or  by 
special  arrangement.  After  having  emerged  from  the  Kondratieff 
depression  in  a  liquid  position,  they  had  a  difficult  time  from  the  second 
half  of  1886  until  the  last  quarter  of  1893  and  the  first  half  of  1894.  Then 
they  recovered  elbowroom,  but  in  their  situation  they  could  do  little  to 
keep  it.  Nor  were  they  able  to  fortify  their  situation  systematically  and 
to  plan  ahead,  as  the  Reichsbank  did  after  1902.  Hence,  after  five  years 
of  easy  steering,  their  surplus  cash  vanished  in  the  course  of  the  events 

1  The  comparison  with  the  graphs  for  outside  banks  is  laborious.     Also,  short-run 
fluctuations  are  much  better  revealed  if  trends  are  eliminated.     For  both  reasons,  the 
reader  is  invited  to  inspect  A.  A.  Young's  charts,  op.  cit.,  pp.  26  and  29. 

2  Since  inflow  of  money  immediately  created  deposits  and  a  corresponding  access  of 
cash  but  not  of  loans,  the  difference  between  deposits  and  loans  must  show  in  the  surplus 
reserve  item.     So  it  does;  see  A.  Piatt  Andrew,  Financial  Diagrams,  No.  22. 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    665 

we  have  glanced  at  in  Chap.  VII.  Responsibility  for  what  no  doubt 
was  a  poor  showing  was  discussed  there.  It  can  not  simply  be  attributed 
to  the  institutional  arrangement.  In  particular,  the  argument  that  this 
arrangement  threw  the  burden  of  cyclical  tensions  on  the  New  York 
banks  and  purchased  comparatively  smooth  working  of  the  outside 
banks  at  the  price  of  unstabilizing  the  former  seems  to  miss  the  point. 
For  this  is  precisely  what  bankers*  banks  are  for.  If  it  be  held  that  they 
were  unequal  to  the  task,  it  should  be  emphasized  that  what  they  were 
unequal  to  was  not  the  handling  of  normal  cyclical  situations  but  the 
prevention  of  those  abnormal  excesses  the  causes  of  which  were  too 
deeply  rooted  in  the  psycho-sociological  pattern  of  the  time  to  be  con- 
trolled by  any  central  bank,  but  the  spirit  of  which  had  infected  them  also. 
The  competitive  character  of  this  system  of  bankers'  banks  was,  how- 
ever, not  its  only  peculiarity.  Perhaps  still  more  important  was  another. 
Owing  to  the  absence  of  an  effective  rediscount  mechanism  and  other 
specifically  American  conditions,  New  York  banks  were  in  particularly 
close  relation  to  stock  exchange  speculation.  Everywhere  member  banks 
lend  to  the  stock  exchange,  and  everywhere  bankers'  banks  thus  indirectly 
helped  to  finance  both  new  issues  and  speculation.  New  York  banks 
would  have  done  so  even  if  they  had  been  pure  member  banks.  But  in 
addition  they  applied  to  this  purpose  by  far  the  greatest  part  of  the  funds 
that  came  to  them  from  their  member  correspondents,  so  that  financing 
the  New  York  stock  exchange  was  the  direct  complement  of  their  bankers' 
bank  activity,  and  stock  exchange  call  loans  stood  in  the  place  which  in 
Europe  was  occupied  by  surplus  cash  or  by  secondary  reserve  items  of  a 
very  different  character.  Often  even  time  loans  were  made  to  serve  the 
purpose.  If  member  banks  lent  surplus  funds  directly  and  used  their 
New  York  correspondents  merely  as  agents — which  was,  as  a  rule,  the 
more  profitable  method — the  situation,  while  substantially  the  same, 
admitted  of  still  less  initiative  on  the  part  of  the  bankers'  banks.  No 
comment  is  necessary;1  but  the  reader  should  allow  himself  to  be  impressed 
by  the  close,  though  not  perfect,2  covariation  of  New  York  Clearings, 
both  with  Value  of  Stock  Exchange  Transactions  and  with  Stock  Prices, 
which  our  chart  displays. 

1  Nor  is  it  necessary  to  comment  on  the  practices  concerning  the  "float"  and  "gross 
deposits"  which  it  would  be  so  enlightening  to  discuss  for  the  purposes  of  a  fuller  diagnosis 
of  the  nature  of  that  situation. 

2  Not  only  do  imperfections  occur  as  to  trends  and  on  many  individual  occasions,  but 
there  is  also  in  the  raw  figures  of  clearings  a  seasonal,  which  is  absent  in  stock  prices.     After 
all,   New  York  industry  and  commerce  was  bound  to  assert   itself.     The  Kondratieff 
branches  show  equally  well  in  all  series — the  "breaks  in  trend" — though  with  unequal 
emphasis.     So  they  do  in  loans  and  deposits.     Juglar  phases  differ.     Value  of  Transac- 
tions rises  from  76,  Clearings  from  78,  for  instance.     Call  rate,  on  the  whole,  agrees  better 
with  Value  of  Transactions  than  with  Clearings. 


666  BUSINESS  CYCLES 

C.  The  Cyclical  Aspects  of  International  Relations. — These  aspects 
of  international  relations  cannot  receive  due  attention  within  this  book, 
and  very  few  remarks  can  be  offered  in  addition  to  what  has  been  said  on 
various  occasions  in  the  historical  chapters.  Of  all  the  limitations 
imposed  by  the  plan  and  purpose  of  this  book,  this  is  the  most  serious 
one.  Not  only  do  cycles  in  different  countries  systematically  affect  each 
other,  so  much  so  that  the  history  of  hardly  any  one  of  them  can  be 
written  without  reference  to  simultaneous  cyclical  phases  in  other  coun- 
tries, but  cycles  really  are,  especially  as  regards  the  great  innovations  that 
produce  the  Kondratieffs,  international  phenomena.  That  is  to  say, 
such  a  process  as  the  railroadization  or  the  electrification  of  the  world 
transcends  the  boundaries  of  individual  countries  in  such  a  way  as  to  be 
more  truly  described  as  one  world-wide  process  than  as  the  sum  of  dis- 
tinct national  ones.  Capitalism  itself  is,  both  in  the  economic  and  in 
the  sociological  sense,  essentially  one  process,  with  the  whole  earth  as  its 
stage.  Both  reasons — interactions  and  supernational  unity  of  funda- 
mental processes — explain  why  in  our  historical  survey  the  cycles  in  our 
three  countries  were  found  to  be  so  much  in  step.1  The  fact  that  they 
were  is  not  more  obvious  than  the  mechanism  that  produced  it  and  also — 
in  principle,  at  least — the  manner  in  which  these  relations  affected  the 
working  of  prewar  central  banks  and  of  the  prewar  gold  standard. 

1.  Even  if  international  economic  relations  consisted  of  nothing  but 
trade  in  commodities  and  services,  the  cyclical  behavior  of  exports  and 
imports,  as  has  been  pointed  out  before,  could  not  be  expected  to  be  as 
regular  as  that  of  other  aggregates.  Not  even  as  to  trend  can  any  general 
proposition  be  formulated,  for  the  process  of  capitalist  evolution  may 
work,  and  at  some  times  and  in  some  countries  actually  has  worked,  in 
the  direction  of  increasing,  instead  of  in  the  direction  of  decreasing,  the 
autarky  of  nations,  quite  independently  of  any  policy  aiming  at  the  former 
end.  In  a  cyclical  movement  fashioned  according  to  our  pure  model 
expectation  would,  if  that  movement  were  confined  to  one  country  and 
if  the  economic  process  in  the  others  were  stationary  or  merely  growing 
(in  our  sense),  be  for  decrease  of  exports  and  increase  of  imports  in  the 
positive  phase,  and  for  the  opposite  behavior  of  both  in  the  negative 
phase.  We  cannot  hope  to  find  this,  of  course.  But  traces  of  it  show 
in  many  instances,  so  for  this  country  from  1872  to  1878,  in  1881  to  1882, 
and  in  1907.  More  convincing  than  totals,  however,  are  imports  and 
exports  of  non-agricultural  commodities.  Hence,  business  situations  in 
the  other  countries  would  in  this  case  tend  to  display  an  inverse  "cycle." 
If  the  innovations  which  are  responsible  for  a  given  cycle  in  a  given 
country  directly  act  on  foreign  countries,  opening  them  up,  for  instance, 

1  On  this,  see  W.  C.  Mitchell,  Business  Cycles,  the  Problem  and  the  Setting,  Chap.  IV, 
Sec.  V. 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    667 

as  markets  in  which  to  buy  or  to  sell,  other  expectations  would,  of  course, 
follow.  If  the  cyclical  process  is  general,  in  the  sense  that  all  the  coun- 
tries that  trade  with  each  other  display  a  cyclical  movement  of  their 
own,  relations  become  much  more  complex.  In  the  simplest  case  in 
which  the  cycles  are  synchronous  and  all  innovations  national,  i.e., 
such  that  they  do  not  directly  affect,  by  rivalry  or  complementarity, 
foreign  industrial  and  commercial  structures,  effects  on  exports  and 
imports  become  a  question  of  relative  intensity  of  phases.  But  if  direct 
interference  with  foreign  industrial  processes  superimposes  itself  on  these 
effects,  we  get  a  rich  tableau  of  possible  cases,  development  of  which 
seems  to  the  writer  to  be  one  of  the  most  important  of  the  reforms  of 
which  the  theory  of  international  trade  stands  in  need.1  The  reader 
should  experiment  with  this  suggestion. 

We  will  confine  ourselves  to  observing  that  although,  in  general,  a 
behavior  of  import  and  export  series  must  be  expected  to  result  from 
this  which  at  first  glance  may  seem  erratic,  yet  conditions  and  relations 
of  any  given  country  are  sufficiently  stable  for  sufficiently  long  periods  to 
make  it  possible  for  us  to  get  along  in  many  cases  with  quite  simple 
theoretical  patterns.  This  stability  shows,  for  example,  in  Professor 
Taussig's  famous  study  on  British  terms  of  trade  by  the  almost  perfect 
inverse  association  of  the  variations  in  the  Board  of  Trade's  wage  index 
with  the  variations  in  net  barter  terms  (equal  to  ratio  of  price  index  of 
imports  to  price  index  of  exports).2 

From  the  standpoint  of  a  central  bank,  that  part  of  international 
commodity  movements  which  comes  about  in  response  to  changes  in 
cyclical  phases  is  in  some  respects  a  corrective  and  in  other  respects  a 
disturbance  of  the  process  with  which  it  has  to  deal.  Any  increase  in 
imports  in  prosperity  and  any  increase  in  exports  in  depression  would  be, 
and  actually  often  has  been,  a  stabilizing — respectively,  restraining  and 
supporting — influence.  But,  in  general,  the  impact  of  foreign,  and  the 
repercussion  of  the  effects  on  foreign  countries  of  domestic,  innovations, 
prosperities,  and  depressions  will  incessantly  affect  terms  of  trade,  quan- 
tities produced  and  in  the  course  of  being  produced,  employment,  credit 
requirements,  and,  if  we  assume  unfettered  gold  standard  all  round,  cash 
items  and  reserves,  in  a  way  which,  whenever  international  relations  are 
important  enough,  currently  disturbs  the  domestic  situation  much  as 
noncyclical  or  noneconomic  disturbances  would.  The  point  to  be  made 
stands  out  still  more  clearly  if  we  assume,  first,  that  the  domestic  banking 
system  also  contributes  to  financing  the  foreign  parties  to  the  trade  or 

1  This  has  been  recognized  and  to  a  considerable  extent  accomplished  by  Mr.  Spencer 
Pollard  in  an  unpublished  manuscript. 

2  The  change  in  Great  Britain's  foreign  trade  terms  after  1900:  Economic  Journal, 
March  1925,  see  chart  on  p.  6.     For  this  country,  see  C.  J.  Bullock,  J.  H.  Williams,  and  A. 
S.  Tucker  in  the  Review  of  Economic  Statistics,  July  1919. 


668  BUSINESS  CYCLES 

else  a  foreign  financial  center  contributes  to  financing  the  domestic  trade, 
and,  second,  that  the  trade  in  every  individual  commodity  or  group  of 
commodities  acquires  an  inertia  of  its  own,  which  in  the  short  run  makes 
it  all  but  independent  of  what  happens  to  the  trade  in  other  commodities 
or  groups  of  commodities.1  Three  things  may  ensue  in  consequence:  the 
domestic  business  situation  may  acquire  a  complexion  completely  at 
variance  with  the  phase  of  the  domestic  cycle — foreign  war  demand  may, 
for  example,  turn  a  depression  into  a  violent  boom;  the  state  of  credit 
may  be  at  variance  with  the  actual  business  situation  prevailing;  and 
the  central  bank  may  be  unable  to  act  in  the  way  in  which  its  diagnosis 
of  the  domestic  situation  would  otherwise  induce  it  to  act — it  may  find 
itself  dependent  on  conditions  in  some  foreign  country.  It  is  at  such 
junctures — which  mere  interaction  of  national  cycles  might  produce — 
that  central  bank  action  is  most  nearly  "initiative"  with  reference  to  the 
domestic  situation. 

2.  But  this  analysis  is  inadequate — and  so,  in  this  respect,  is  the 
general  theory  of  international  trade — because  it  bases  international 
relations  on  commodity  trade,  which  in  turn  harks  back  to  primitive 
barter  and  to  which  financial  transactions  are  in  principle  ancillary. 
We  need  not  go  into  the  question  how  great  the  sector  of  reality  is  or  was 
for  which  this  model  could  be  considered  to  be  satisfactory.  For  the 
great  mass  of  transactions  that  make  up  the  commerce  of  nations  in  the 
capitalist  epoch  it  is  clear,  however,  not  only  that  the  stage  has  at  any 
given  time  been  set  for  them  by  the  cyclical  process  of  evolution,  but  also 
that  priority  in  the  mechanism  of  economic  relations  between  nations 
belongs  not  to  trading  but  to  finance.  It  would  be  truer  to  say  that 
modern  commodity  trade  followed  and  complemented  capital  trans- 
actions than  that  the  latter  arose  out  of  and  complemented  commodity 
trade.  Selling  presupposes  lending  or  "capital  export"  in  other  forms, 
and  commerce  develops  within  environments  first  created  and  incessantly 
reshaped  by  entrepreneurial  and  capitalist  ventures. 

For  our  purpose  it  is  sufficient  to  consider  the  effects  of  capital  move- 
ments on  the  cyclical  situations  and  policies  of  banks,  especially  central 
banks.  We  distinguish  long-term  and  short-term  transactions  and  again 
transactions  which  arise  from  the  business  sphere — hence  mainly  from  our 
process — and  transactions  which  do  not — mainly  public  borrowings,  which 

1  This  is  the  element  of  truth  in  the  statement  that  commodity  trade,  even  if  the  only 
form  of  international  economic  intercourse,  need  not  balance,  because  it  depends  on  the 
wants  of  the  trading  countries  for  one  another's  products.  No  defense  of  this  statement, 
which  in  most  cases  is  indicative  of  a  failure  to  grasp  the  elements  of  the  theory  of  inter- 
national trade,  is  intended  here.  But  if  we  make  the  assumptions  that  are  necessary  in 
order  to  enable  us  to  accept  the  classical  description  of  an  equilibrating  mechanism  acting 
on  incomes  and  prices  through  gold  movements,  we  readily  perceive  the  possibility  that 
its  working  may  produce  an  otherwise  unmotivated  slump  or  boom. 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    669 

we  assume  to  be  independent  of  the  cyclical  process,  although  this  is  no 
more  than  an  expository  simplification — and  which  hence  impinge  on 
money  and  open  markets  at  random.  Suppose  that  a  banking  house, 
say,  in  the  London  of  the  late  nineties,  floated  a  bond  issue  which  was  to 
finance,  for  example,  electrical  enterprise  in  Argentina,  that  subscriptions 
were  partly  financed  by  bank  loans,  and  that — in  order  to  relieve  our 
argument  from  a  few  obvious  propositions — the  proceeds  were  intended  to 
be  spent  in  the  borrowing  country.  The  borrower  acquired  a  balance 
in  London  and  could,  if  he  had  wished  to,  have  acquired  the  balance  in 
Argentina — which  was,  under  our  assumptions,  what  he  really  wanted — 
by  converting  the  English  balance  into  notes  of  the  Bank  of  England  and 
these  into  sovereigns  which  he  might  have  shipped  and  deposited  with  his 
Argentine  Bank.  Further  assuming  that  the  transaction  was  big 
enough,  his  spending  would  have  imparted  an  impulse  to  the  business 
situation  in  Argentina  which,  in  turn,  would  have  supplied  the  reason  why, 
on  its  access  of  cash,  the  Argentine  bank  would  have  been  not  only  willing 
but  also  able  to  expand  its  loans,  and  so  on.  But  in  England  such  a 
transaction,  or  a  sufficiently  important  bundle  of  such  transactions, 
might  have  brought  down  the  credit  structure  by  suddenly  enforcing 
violent  contraction.  It  is  true  that  foreign  borrowers  could  not  have 
hoped  for  attractive  conditions,  or  issuing  houses  for  success,  in  a  situa- 
tion that  was  already  tight  for  other  reasons.  It  is  also  true  that  member 
banks,  since  they  were  normally  not  loaned  up,  and  the  Bank  of  England, 
which  always  kept  surplus  funds,  would,  in  general,  have  been  able  to 
mitigate  the  shock.  In  reality,  moreover,  our  borrower  would  not  as  a 
rule  have  wished  to  spend  all  the  proceeds  of  the  loan  at  home.  Even 
if  he  had,  he  would  not  have  drawn  gold  as  assumed,  but  would  gradually 
have  sold  exchange  on  London.  Equilibrating  effects  on  the  preexisting 
commodity  trade  would  eventually  have  worked  themselves  out  through 
changes  in  incomes  or  in  incomes  and  prices.  The  preexisting  commodity 
trade  itself  and  its  financial  complement  always  contained  sloppy  nooks 
and  crannies.  Assets  which  the  lending  country  owned  in  the  borrowing 
or  other  countries  could  have  been  drawn  upon  to  some  extent.  Thus, 
the  process  worked  on  lines  all  of  which  were  studded  with  additional 
shock  absorbers. 

But  the  facts  will  remain,  first,  that  during  a  period  of  indefinite 
length  a  net  disturbing  influence  on  the  lending  country  would  almost 
unavoidably  have  been  exerted  by  such  transfers  on  capital  account, 
since  there  was  no  equilibrating  mechanism  that  would  have  been  fully 
effective,  except  in  the  long  run ;  and,  second,  that  the  disturbance  was  in 
this  case  due  to  the  international  gold  standard — which,  contrary  to 
fact,  we  assumed  also  for  Argentina — for  without  it  Argentina  might 
have  done  just  as  well  with  a  corresponding  volume  of  domestic  "infla- 


670  BUSINESS  CYCLES 

tion"  as  she  did  with  the  gold.  Hence,  the  gold  standard,  whatever  its 
merits  or  demerits  may  be  in  other  respects,  here  created  a  problem  by 
throwing  on  the  central  banks  of  gold  countries  a  burden  that  did  not 
arise  from  but  interfered  with  their  own  economic  process.  No  great 
importance  is  claimed  for  that  type  of  international  transactions;  but  an 
element  of  what  the  example  was  intended  to  stress  entered  into  almost 
any,  and  the  consequent  situations  enforced  much  more  initiative  action 
by  the  central  banks  than  any  that  could  conceivably^have  arisen  from 
pure  commodity  trade.  We  may  now  add  issues  of  foreign  governments1 
— which  in  the  time  of  economic  liberalism  were  not  always  dependent 
on  the  consent  of  either  the  central  bank  or  the  foreign  office  of  the  lending 
country,  and  which  sometimes  created  balances  that  put  the  debtor 
temporarily  in  a  position  of  considerable  power — and  repercussions  of 
foreign  disturbances  of  both  economic  and  extraeconomic  nature.  It 
then  becomes  still  clearer  that  a  capitalist  country's  financial  and  business 
situation  was  at  any  time  the  result  of  a  national  (cyclical)  and  of  an 
international  component,  which  were  largely  independent  of  each  other 
as  to  causation  and,  hence,  likely  to  combine  in  an  erratic  way. 

The  very  logic  of  a  central  bank's  position  and  its  own  interest  made 
it  imperative  to  try  to  coordinate  the  two,  i.e.,  to  protect  the  domestic 
process  from  being  upset  by  international  transactions  and  disturbances 
without  impairing  the  foreign  business  of  the  nation,  and  in  particular  to 
manage,  if  possible  in  advance,  those  states  of  liquidity  and  stringency  in 
the  market  which  under  the  circumstances  were  not  infrequently  out 
of  step  with  the  phase  of  the  domestic  cycle.  It  is  but  commonplace 
to  say  that  the  endeavor  to  do  this — under  the  conditions  set  by  the 
gold  standard — became  the  chief  motive  of  central  bank  policy,  in 
England  understandably  more  than  anywhere  else,  and  that  the  struggle 
of  central  banks  for  control  over  the  money  market  (and  for  the  freedom 
of  action  such  control  would  entail)  was  also  chiefly  motivated  by  it. 
This  commonplace  is,  however,  not  without  diagnostic  value,  in  view 
of  the  emphasis  put  by  so  many  economists  on  the  role  that  central  banks 
do  (or  ought  to)  play  in  the  cycle  per  se.  This  is,  as  we  have  seen,  not 
the  crux  of  the  matter.  But  owing  to  the  presence  of  that  other  com- 
ponent, monetary  management  of  an  entirely  different  type  and  of  much 
larger  scope  imposed  itself  nevertheless — in  the  heyday  of  laisserfaire — 
and  would  have  imposed  itself,  though  to  a  lesser  extent,  even  without 
the  strain  put  on  the  working  of  the  gold  standard  by  what  we  have  called 
(Chap.  VII,  Sec.  E)  neomercantilist  policies.  Moreover,  it  was  this 

1  In  markets  smaller  than  the  London  market  the  current  operations  of  foreign  govern- 
ments arising  both  from  their  long-term  financing  and  from  the  trade  in  commodities  were 
sometimes  a  source  of  major  trouble.  The  operations  of  the  Russian  government  in  Berlin 
during  the  tenure  of  office  of  Wyshnegradsky  and  Witte  afford  instances. 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    671 

component  that  made  gold  movements  so  important  to  central  banks. 
The  inflows  and  outflows  attributable  to  the  cyclical  component  would, 
in  themselves,  have  been  a  very  minor  consideration.  But  gold  move- 
ments induced  by  international  capital  transactions  were  not  only  symp- 
toms of  a  lack  of  coordination  of  the  two  components  but  possibly 
dangerous  in  themselves.  They  so  regularly  called  for  action  that  it  was 
easy  to  draw  the  conclusion  that  all  the  central  banks  minded  was  their 
reserve. 

But  the  art  of  central  banking  precisely  consisted  in  minding  the 
nation's  business  and  in  so  steering  between  the  possibilities  of  catastrophe 
in  domestic  affairs  and  of  discomfiture  in  international  affairs  that  the 
compromise  would  be  tolerable  to  both.  Considering  that,  owing  to 
England's  huge  capital  exports  and  international  interests  in  general, 
the  smallest  mistake  could  at  times  have  produced  disastrous  conse- 
quences, actual  performance,  working  with  so  small  a  gold  reserve,  was 
truly  remarkable.  The  resulting  irregularities  in  the  behavior  of  foreign 
exchanges  and  gold  movements,  neither  of  which  displays  consistent 
cyclical  patterns,  cannot  be  here  discussed.  A  few  remarks  on  various 
procedures  chosen  by  the  Bank  of  England  will,  however,  suffice  to  give 
a  rough  idea  of  the  extent  to  which  we  must  be  prepared  to  find  deviations 
from  expectation  based  upon  our  model  in  monetary  and  banking  series. 
It  is  clear  from  the  outset  that  these  deviations  cannot,  except  tem- 
porarily, have  been  very  important,  because  otherwise  our  interpretation 
of  those  series  by  means  of  that  model  alone  could  not  have  accounted 
for  general  contours  as  it  did.  This  in  itself  testifies  to  the  success  of  the 
Bank.  For  it  implies  that  the  Bank  succeeded  in  managing  external  dis- 
turbance while  retaining  freedom  to  act  with  respect  to  domestic  situa- 
tions according  to  the  principles  discussed  in  the  preceding  section. 

In  the  first  place,  it  has  to  be  recognized  that  what  foreign  financing 
primarily  imposed  was  increased  caution  and  the  necessity  of  preventing 
liquidity  from  engendering  tightness.  This,  together  with  the  fact  that 
in  a  system,  which  works  its  funds  so  scientifically,  open-market  rates 
will,  whenever  the  situation  of  the  moment  permits  it,  immediately  tend 
to  fall  to  a  minimum,  explains  why  the  action  of  the  Bank  was  almost 
always  in  the  direction  of  steadying  or  raising  open-market  rates  and  why 
controlling  the  open  market  came  to  mean  tightening  it.  This  did  not 
necessarily  mean  that  it  wished  to  make  money  dearer  for  domestic 
business.  On  the  contrary,  measures  were  sometimes  taken  to  avoid  that 
when  action  on  the  latter  was  not  cyclically  indicated,  as,  for  instance, 
during  the  crisis  of  1907  and  in  the  notable  case  of  1910. l  Tightening  the 
1  The  simplest  way  to  acquire  such  grasp  of  the  matter  as  can  be  acquired  from  reading, 
is  to  peruse  the  volumes  of  the  Economist  and  the  Statist.  But  the  comments  offered  in 
both,  especially  the  former,  however  sound  from  the  standpoint  of  a  world  that  was  resolved 
to  play  the  capitalist  game,  should  not  be  uncritically  accepted. 


672  BUSINESS  CYCLES 

open  market  without  tightening  the  money  market  (in  our  sense),  or 
adjusting  liquidity  to  the  ruling  cyclical  phase,  may,  in  fact,  serve  as  a 
formula  by  which  to  express  one  type  of  those  moves  by  which  the  Bank 
of  England — and  similarly  also  the  Reichsbank — attempted  to  coordinate 
the  two  components  distinguished  above.  Undesired  pressure  on  domes- 
tic business  was  no  doubt  exerted  all  the  same,  but  to  an  extent  only 
which  in  the  light  of  the  analysis  of  Chaps.  XI  and  XII  and  of  the 
preceding  section  cannot  be  held  to  have  produced  any  major  effects. 

In  the  second  place,  it  is  easy  to  see  why,  as  far  as  central  bank  policy 
aimed  at  coordination  in  this  sense,  measures  as  to  bank  rate1  were 
distinctly  secondary  to  open-market  operations  and  to  a  peculiar  kind  of 
suasion.  The  Bank  borrowed  in  the  market — or  from  the  London  banks 
directly  so  that,  their  loans  to  the  market  being  curtailed,  the  latter  was 
driven  into  the  Bank2 — or  sold  spot  and  repurchased  on  account,  or 
sold — much  more  rarely  bought — outright,  primarily  in  response  to  con- 
ditions that  arose  from  the  international  business,  in  order  to  be  able  to 
deal  with  the  domestic  situation  on  its  merits,  although,  of  course,  in 
many  cases  both  classes  of  considerations  pointed  in  the  same  direction. 
The  peculiar  kind  of  suasion  consisted  in  securing  the  cooperation  of 
individual  holders  of  big  balances,  such  as  India  House  or,  for  a  time,  the 
Japanese  government.  In  the  paradigma  above  discussed  one  of  the  first 
things  which  it  would  have  occurred  to  the  Bank  to  do  would  have  been 
to  convey  to  the  Argentinian  concern  the  impression  that  the  Bank  had 
plenty  of  means  to  make  things  in  the  future  more  or  less  comfortable 
for  it,  and  that  it  was  just  as  well  not  to  withdraw  gold  in  an  inconvenient 
way  or  to  create,  by  lending  in  the  open  market,  a  state  of  liquidity  the 
Bank  did  not  wish  to  see.  The  gratitude  of  the  Bank  was  an  asset  that 
it  may  have  paid  to  acquire  at  the  price  of  some  sacrifice  of  interest. 

In  the  third  place,  if  the  Bank  merely  desired  to  strengthen  its  own 
position  without  interfering  with  domestic  trade  at  all,  the  device  of 
varying  the  purchasing  price  of  gold  was  resorted  to  with  considerable 
success.  Finally,  there  was  the  possibility  of  special  arrangements  with 
foreign  central  banks,  particularly  the  Bank  of  France.  A  well-authenti- 
cated instance  is  afforded  by  the  two  transactions,  the  one  with  the 
Russian  Bank  (i.e.,  the  Russian  government),  the  other  with  the  Bank  of 

1  Those  measures,  in  turn,  consisted  in  manipulating  not  so  much  the  official  rate  as  the 
rate  on  advances  or  in  rediscounting  at  a  rate  higher  than  the  official  one  and  in  similar 
devices.     All  of  them  amounted  to  discrimination  with  respect  to  purpose,  which  on  a  few 
occasions  at  least  was  also  resorted  to  directly.     Cases  of  the  last-mentioned  type  illustrate 
our  argument  particularly  well.     For  it  was  foreign  paper — American,  especially — or  paper 
serving  foreign  purposes,  that  was  discriminated  against. 

2  This,  for  example,  was  done  in  December  1905,  when  it  was  intended  to  put  on  brakes 
also  in  the  domestic  business;  see  the  Economist's  annual  survey  of  1905,  where  it  is,  how- 
ever, stated — incorrectly,  as  the  writer  thinks — that  this  was  a  "new  departure." 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    673 

France,  during  the  Baring  crisis.  Others  are  more  doubtful.1  For  us 
only  two  things  matter:  never  was  such  a  measure  actually  taken  or 
"suspected"  except  in  connection  with  the  international  component  of 
English  situations,  and  in  each  case  the  motive  was  to  protect  the  working 
of  the  domestic  business  organism  from  being  disturbed  by  foreign  dis- 
turbance or  by  repercussions  of  foreign  disturbance  on  English  finance, 
which  is  the  reason  why  our  series  were  not  more  affected;  and  there  is  no 
justification,  either  for  the  patriotic  irritation  displayed  by  the  Economist 
and  some  English  writers  at  the  suggestion  that  that  measure  was  taken 
several  times,  or  for  the  opinion  of  other  writers  that  it  each  time  spelled 
a  breakdown  of  the  prewar  bar) king  policy.  In  the  internationalized 
world  of  the  free  gold  standard,  borrowing  by  one  central  bank  from 
another  was  a  perfectly  natural  thing  to  do  and  nothing  to  be  ashamed  of. 
Management  of  the  international  component,  however,  was  so  success- 
ful only  because  there  was  the  powerful  wall  of  short  and  semiliquid 
claims  on  foreign  debtors  that  sheltered  the  English  structure.  Without 
it,  that  structure  could  never  have  been  worked  on  such  small  margins  of 
safety  or  by  means  of  such  delicate  adjustments.  This  mass  of  claims — 
Mr.  Hartley  Withers  estimated  it  in  1909  at  from  150  to  200  million 
pounds — which  was  currently  turned  into  cash  to  be  presently  reinvested 
almost  anywhere  within  the  gold  area,  responded  to  the  Bank's  slightest 
move  very  much  more  promptly  than  foreign-owned  balances  would  have 
done,  facilitated  great  capital  transactions,  supported  foreign  business, 
mitigated  domestic  stringencies.  Because  of  its  presence,  tightening  the 
open  market — raising  open-market  rates — not  only  regulated  but,  by 
drawing  gold,  eased  situations.  Similarly,  cyclical  increase  in  bank  rate 
had  not  only  restrictive  but  also  relieving  effects,2  and  incidentally — it  is 
quite  misleading  to  make  this  the  center  of  the  matter — turned  unfavor- 
able exchanges,  as  a  rule,  with  the  utmost  ease.  Hence,  we  cannot  be 
surprised  to  find  that  a  general  belief  in  the  effectiveness  of  bank  rate 
grew  up  among  bankers  and  economists  which,  without  this  short  capital 
situation,  it  would  be  difficult  to  understand.  It  is  more  astonishing  that 

1  Explicit  arrangements,  however,  were  not  always  necessary.     It  is  easy  to  see  that, 
in  some  cases  in  which  "assistance"  might  have  been  thought  desirable,  it  would  at  the 
same  time  have  been  to  the  interest  of  foreign  bankers'  banks  to  behave  in  a  way  that  would 
amount  to  a  timely  redistribution  of  gold  reserves.     The  Bank  of  France,  with  its  strong 
aversion  to  variations  in  its  rate,  had  a  particular  motive  for  it.     Whether  the  "help"  was 
extended  to  the  Bank  of  England  or  to  the  market  is  immaterial.     English  paper  was 
actually  taken,  and  gold  released,  by  the  Bank  of  France  in  1906  and  1907,  and  this  eased 
the  situation  and  prevented  further  rise  in  bank  rate.     Whether  this  was  done  by  arrange- 
ment or  not  is  not  particularly  interesting. 

2  The  latter  were  probably  more  important  than  the  former.     Though  the  danger- 
signal  effect  must  be  borne  in  mind,  the  Bank  had,  as  we  have  seen,  not  much  to  do  with 
stopping  normal  booms.     The  classical  theory  of  bank  rate  exaggerated  the  importance  to 
general  business  of  moderate  variations  in  rate  as  much  as  more  modern  theories  do. 


674  BUSINESS  CYCLES 

so  many  people  failed  to  see  the  dependence  of  that  effectiveness  on  the 
historically  unique  technical  position  of  the  London  market  and,  instead, 
tried  to  explain  it  by  a  perfectly  general — and  unrealistic — theory.  This 
theory,  on  the  one  hand,  greatly  overstressed  the  influences  which  bank 
rate  can  exert  on  foreign  exchanges  and  gold  movements  through  suc- 
cessive effects  on  volume  of  domestic  transactions,  volume  of  deposits, 
price  level  and  incomes,  balance  of  international  commodity  trade.  On 
the  other  hand,  it  never  succeeded  in  properly  defining  the  places  that 
open-market  rates  and  reserve  proportions  hold  relatively  to  each  other 
in  the  rational  schema  of  central  bank  policy. 

The  presence  of  that  light  cavalry  of  English  finance  then  explains 
why  both  bank  rate  and  gold  movements  proved,  comparatively  speaking, 
so  little  disturbing  to  business  and  incidentally  to  the  course  of  cyclical 
phases.  As  regards  foreign  exhanges  in  particular,  a  cyclical  regularity 
was  established  by  its  action,  which  should  be  noted  in  passing.  In  the 
absence  of  any  chances,  risks,  and  costs  incident  to  the  transfer  of  short 
balances,  the  short-capital  mechanism  would  obviously  have  tended  to 
equalize  comparable  open-market  rates  in  different  countries.  As 
between  countries  of  different  monetary  standards,  the  element  of  risk 
and  chance  was  dominant.  But  as  between  gold-standard  countries, 
variations  of  exchange  rates  were  confined  within  gold  points.  Hence, 
it  was  possible  for  the  operating  banker  to  calculate  maximum  risk  and 
chance  and  to  compare  the  result  with  the  given  difference  between 
comparable  rates  ruling  in  different  countries.1  Given  the  responsiveness 
of  the  short-capital  mechanism  to  fractional  gains,  this  would  not  only 
limit  possible  deviations  from  equality  of  market  rates,2  but  also  tend  to 
enforce  parallelism  of  variations  in  those  differences  and  variations  in 
rates  of  exchange.  This  expectation  is  disappointed  as  regards  the  rela- 
tion between  London  and  New  York,  but  comes  out  very  well  as  regards 
the  relation  between  London  and  the  European  gold-standard  countries, 
especially  Germany  and  France.3  Still  we  cannot  expect  that  parallelism 

1  There  were,  however,  several  difficulties  about  doing  this.     We  are  assuming  that 
there  were  no  risks  other  than  those  implied  by  the  possible  range  of  variation  in  rate  of 
exchange.     If  we  take  the  highest  grade  bills  only,  this  assumption  works  under  normal 
conditions.     But  it  fails  in  any  really  serious  crisis  and  also  in  the  presence  of  political 
fears.     Moreover,  gold  points  were  somewhat  variable,  both  in  time  and  as  between  banks, 
some  of  which  may  have  had  opportunities  to  ship  more  cheaply  than  others. 

2  As  far  as  the  writer  knows,  this  "solidarity"  between  the  open  markets  of  gold  coun- 
tries was  first  investigated  by  N.  E.  Weill,  Die  Solidaritat  der  Geldmarkte,  1903,  though 
the  main  facts  about  it  were,  of  course,  familiar  to  all  writers  on  banking  and  finance. 

3  See  the  standard  work  on  the  subject,  H.  Neisser,  Der  Internationale  Geldmarkt  vor 
und  nach  dem  Kriege,  Weltwirtschaftliches  Archiv,  April  1929,  especially  charts  on  pp.  187 
and  189,  also,  for  what  follows  the  chart  on  p.  221,  which  compares  the  curve  of  the  distance 
between  the  London-Berlin  exchange  and  the  difference  of  market  rates  with  the  Thomas 
index  of  English  business  conditions. 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    675 

to  be  perfect.  There  are  transactions  that  are  less  sensitive,  or  even  not 
at  all  sensitive,  to  those  differences.  We  have  to  do,  moreover,  with  a 
surface  mechanism  which  cannot  always  control  the  influence  of  the  more 
fundamental  factors  behind  it.  Finally,  anticipations  as  to  the  behavior 
of  exchanges  in  the  relevant  future  will  work  differently  on  different 
levels  as  well  as  in  different  seasonal  and  cyclical  situations.  It  has,  in 
fact,  been  shown  by  Professor  Neisser  that  the  graph  of  the  distance 
between  the  curves  of  exchange  rates  and  of  the  difference  between  market 
rates  displays  fairly  regular  cyclical  properties,  which  are  what  emerges 
behind  the  apparently  erratic  behavior  of  exchange  rates  considered  by 
themselves.  It  does  not  follow  that  we  should  expect  any  consistent 
short-run  relation  between  foreign  exchanges  and  domestic  market  rates 
alone,  or  that  we  should  be  able  to  predict  any  definite  effect  from  a  given 
absolute  change  in  domestic  bank  rate  alone,  however  completely  the 
Bank  may  succeed  in  making  it  effective.  The  opinion  which  it  is  said 
prevailed  in  the  city  and  in  support  of  which  Goschen's  authority  might 
be  invoked,  viz.,  that  an  increase  of  1  per  cent  in  bank  rate  would,  and 
a  smaller  increase  would  not,  turn  unfavorable  exchanges,  was  therefore 
not  exact. 

3.  In  itself,  the  mere  fact  that  there  was  in  the  world  of  prewar 
capitalism  one  gold  market  which  in  importance  and  accessibility  over- 
shadowed all  others1  would  not  greatly  matter  for  our  purpose.  It  would 
merely  offer  an  occasion  for  the  remark  that,  first,  because  of  the  "autono- 
mous" movement  of  gold  from  the  countries  which  produced  it,  and 
second,  because  of  the  special  transactions  effected  by  central  banks  or 
other  monetary  authorities,  no  expectation  as  to  cyclical  behavior  can  be 
formed  with  any  confidence.  The  cyclical  migrations,  which  the  units 
of  a  constant  stock  of  gold  could  have  been  expected  to  display  under  a 
regime  of  unfettered  gold  currency  and  in  the  absence  of  any  regulatory 
interference,  are  too  clear  to  detain  us.2  But  all  the  more  relevant  is  the 
further  fact  that  the  gold  market,  toward  which,  with  the  partial  excep- 
tion of  the  production  of  the  United  States  and  Russia,  practically  all 
newly  produced  gold  gravitated,  was  located  in  the  same  country  and 
controlled  by  the  same  financial  organism  that  also  controlled  the  inter- 
national short-loan  fund.  Because  of  this,  England  was  in  a  position,  by 
extending  or  restricting  short  credits,  to  enable  operators  in  other  coun- 
tries to  buy  gold  or  else  to  prevent  them  from  doing  so,  not  absolutely 

1  There  were  many  local  ones  all  over  the  world.     In  Paris  a  price  was  regularly  quoted 
for  bars  which  was  distinct  from  the  gold  premium  charged  by  the  Bank  and  from  the  price 
of  gold  exchange. 

2  This  does  not  mean  that  they  do  not  offer  any  problems,  but  only  that  the  problems 
relevant  to  our  argument  can  readily  be  solved  if  we  amend  the  classical  theory  of  gold 
movements  by  a  theory  of  the  short-balances  mechanism  and  stress  differences  in  interest 
rates  rather  than  differences  in  pricelevels. 


676  BUSINESS  CYCLES 

but  to  a  considerable  extent.  As  far  as  the  Bank  of  England  controlled 
the  domestic  open  market,  it  may,  therefore,  be  said  to  have  controlled 
international  gold  movements  and  to  have — indirectly — acted  as  the 
bankers'  bank  of  bankers'  banks.  This  strengthened  its  position  and  its 
ability  to  mitigate  the  impact  of  foreign  disturbances  on  domestic  busi- 
ness situations  in  a  way  not  open  to  any  other  central  banks.  The 
Reichsbank,  the  Bank  of  France,  and  other  central  institutions  aimed  at 
similar  freedom  of  action  by  other  means.  Some  of v  these  were  also 
adopted  by  the  Bank  of  England.  It  was,  in  fact,  a  very  natural  dis- 
covery to  make,  that  actual  or  expected  gold  movements  might  be 
influenced  directly  and  that  this  method  would  serve  to  handle  a  class  of 
short -run  difficulties  as  effectively  as  measures  which,  besides  doing  this, 
would  also  produce  undesired  effects.  But  while  other  central  banks 
began  at  the  turn  of  the  century  to  head  toward  the  gold-exchange  stand- 
ard, which  meant,  though  they  may  not  have  been  aware  of  it,  moving 
away  from  gold-standard  ideas,  the  Bank  of  England,  up  to  the  war,  never 
did  anything1  that  pointed  in  that  direction.  All  it  attempted  to  do  was 
to  smooth  the  working  of  the  unfettered  gold  standard,  which  for  the 
time  being  was  clearly  in  the  interest  of  England  and  which  remained  her 
long-run  policy.  It,  therefore,  confined  itself  mainly  to  acting  on  gold 
points  by  varying  its  price  for  bars  and  foreign  coins,  refusing  to  sell  bars, 
giving  free  advances  on  gold  imports,  and  so  on,  and  in  all  normal  cases 
continued  to  rely  on  its  influence  on  money  rates.  To  insulate  the 
domestic  price  level  was  no  part  of  this  policy.  To  stabilize  it  was  part 
of  the  policy  only  in  the  sense  that — never  mind  what  the  Bank's  leading 
men  may  have  thought  or  said — it  tended  to  mitigate  temporarily  the 
impact  of  noncyclical  factors. 

In  the  late  eighties  and  the  early  nineties — from  1879  to  1888,  total 
coin  and  bullion  in  the  Bank  of  England  fell  (with  fluctuations)  by  about 
one-third,  while  total  deposits  in  the  United  Kingdom  rose  by,  roughly, 
one-third — the  problem  was  one  of  "scarcity."  It  was  during  that  time, 
as  far  as  the  writer  knows,  that  the  Bank  first  resorted  to  gold  policies, 
which,  however,  proved  particularly  effective  during  the  troubles  of  the 
early  nineties.2  What  would  have  happened  if  no  increase  in  gold  pro- 
duction had  occurred  is  easy  to  guess:  new  methods  of  economizing  gold 
would  have  been  resorted  to.  As  it  was,  South  African  gold  reversed  the 
situation.  Its  modus  operandi  is  highly  instructive.  The  Bank  at  first 
simply  discontinued  its  gold  policy  and  allowed  the  new  gold  to  accumu- 
late in  its  vaults — its  gold  stock  increased  rapidly  from  1891  to  1895,  and 

1  Fulfilling  its  legal  obligation  to  sell  sovereigns  in  minimum-weight  coins  was  as  near 
as  it  went  to  refusing  to  play  the  orthodox  game. 

2  In  1893,  when  the  inflow  of  gold  had  already  set  in,  the  price  of  gold  in  the  London 
market  rose  to  8  pounds,  17  shillings,  10.8  pence  per  ounce  standard,  while  bank  rate 
stopped  at  4  per  cent. 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    677 

its  reserve  proportion  moved,  substantially,  parallel  to  it1 — or  to  flow 
through  the  London  market  into  other  banks2  and  to  other  countries, 
while  the  price  level  continued  to  fall.  The  effect  on  open -market  rates  is 
as  well  marked  as  we  should  expect,  and  was  temporarily  intensified  by 
the  payment  to  Japan  of  the  Chinese  indemnity,  which  involved  the 
transfer  to  London  in  October  and  November  1895  of  13  million  pounds 
raised  by  the  Chinese  government  in  Paris.  But  it  should  be  obvious 
that  even  this  effect  of  the  new  gold  cannot  be  interpreted  in  terms  of 
monetary  factors  alone  and  that  it  was  so  much  in  evidence  only  because 
of  the  cyclical  situations  on  which  the  new  gold  happened  to  impinge. 
For,  in  the  prosperity  that  followed,  it  entirely  disappeared,  and  rates 
stiffened  in  spite  of  the  swelling  tide  of  gold  production.  Prices,  on  the 
other  hand,  then  began  to  rise.  They  probably  rose  more  than  they 
would  have  done  in  the  absence  of  such  a  tide,  exactly  as  money  rates 
had  no  doubt  fallen  more  than  they  would  have  done  in  the  absence  of  it. 
Neither  effect  is  being  denied  here,  but  neither  was  strong  enough  to 
assert  itself  except  in  cyclical  phases  favorable  to  it.  The  writer  does  not 
see  that  this  interpretation  leaves  any  major  fact  unaccounted  for  or 
that  there  is  in  the  sequence  of  these  events  anything  to  contradict  our 
views  about  the  cyclical  process,  its  relation  to  monetary  factors,  and  the 
facts  and  possibilities  of  central  bank  policy. 

To  complete  this  part  of  our  argument,  we  will  notice  that  even  before 
prosperity  (in  our  sense)  had  set  in  and  while  revival  (in  our  sense)  was 
in  full  swing,  with  low  rates,  low  prices,  and  great  activity  in  building 
and  in  trade  in  general,  the  American  troubles  of  1896  and  the  incident 
drain  of  gold  drove  the  Bank  back  upon  raising  its  selling  price  for  bars. 
It  also  raised  its  rate,  which,  in  a  very  liquid  market,  it  proved  extremely 
difficult  to  make  effective.  The  latter  measure,  not  easy  to  understand 
from  the  situation  in  which  it  was  taken,  was  ratified  by  events,  for  that 

1  See  W.  E.  Beach,  British  International  Gold  Movements  and  Banking  Policy,  1881- 
1913,  1935,  p.  71. 

2  This  was  then,  but  especially  later,  facilitated  by  what  may  be  called  the  Gold  Scare. 
Ever  since  the  publication  of  Bagehot's  Lombard  Street,  financial  writers  had  been  preach- 
ing that  the  "gold  basis"  of  the  British  banking  system  was  too  small.     These  preachings 
began  to  take  effect  in  the  late  nineties,  and  the  great  London  banks  used  the  opportunity 
afforded  by  the  new  plethora  of  gold  to  build  up  gold  reserves  of  their  own.     Whatever 
we  may  think  of  the  sentiments  voiced  in  Sir  Felix  Schuster's  addresses,  for  example,  and 
of  the  reality  of  the  "dangers"  incident  to  smallness  of  the  "ultimate  reserve,"  the  scare 
certainly  helped  to  produce,  together  with  the  demand  for  gold  from  the  countries  which 
at  that  time  went  on  the  gold  standard  or  made  preparations  for  doing  so  or  simply  wished 
to  broaden  the  basis  of  an  already  existing  gold  currency,  much  the  same  results  as  could 
have  been  expected  from  an  international  policy  of  partial  sterilization  of  the  new  gold. 
This  is  one  of  the  reasons  why  it  is  not  easy,  on  the  one  hand,  to  interpret  the  gold  move- 
ments of  that  period  and,  on  the  other  hand,  to  estimate  the  net  effect  of  the  new  gold  on 
price  levels. 


678  BUSINESS  CYCLES 

situation  shaded  off  into  a  burst  of  prosperity  the  following  year  when 
the  price  of  bars  was  raised  to  £  3-18-J^,  but  bank  rate  to  not  more  than 
3  per  cent.  The  course  of  events  in  America  and  Germany,  the  require- 
ments of  foreign  financing  and  boom  conditions  at  home,  account  for  the 
use  of  gold  devices  along  with  bankrate  in  the  next  two  years,  during 
which  the  Bank's  gold  stock  fell  to  the  level  of  1894.  Again,  American 
troubles  and  war  finance  account  for  their  use  instead  of  additional 
manipulations  of  bank  rate,  from  1900  on.  As  the  Kendratieff  wore  on 
and  gold  continued  to  flow  in,  the  Bank's  position  grew  stronger  of 
itself;  and  toward  the  end  of  the  period,  protective  gold  policies  fell 
almost  into  disuse.  There  are,  in  fact,  traces  of  a  preoccupation  with  the 
opposite  problem,  i.e.,  the  problem  of  preventing  the  gold  stock  of  the 
Bank  from  increasing  to  amounts  the  very  sight  of  which  would  have 
encouraged  excesses.  It  was,  however,  kept,  in  its  yearly  averages,  at  or 
near  the  30  million  level  until  1913. 

No  such  gold  management  was,  of  course,  possible  in  the  United 
States,  where,  in  consequence,  we  find  during  the  sixteen  years  preceding 
the  war  much  more  parallelism  between  national  bank  deposits  and  the 
total  monetary  gold  stock.  The  German  case  is  for  our  purpose  less 
interesting.  The  Reichsbank  was  mainly  concerned  with  strengthening 
the  bases  of  the  German  standard.  It  was  a  steady  buyer  of  gold  from 
the  first,  acquiring  over  434  million  dollars  of  it  between  1876  and  1893. 
In  the  times  of  prosperity  that  followed,  such  gains  as  were  made  were 
more  than  absorbed  by  circulation.  But  from  1907  on,  the  accumulation 
of  a  larger  reserve  tended  to  become  an  end  in  itself,  and  all  other  con- 
siderations were  subordinated  to  it.  By  purchases  and  the  use  of  gold- 
exchange  standard  and  other  devices,  among  which  the  substitution  of 
the  Reichskassenscheine  for  gold  in  circulation  may  be  mentioned,  it  suc- 
ceeded in  acquiring  the  ample  gold  outfit  with  which  it  entered  the  war. 

D.  Stock  Exchange  Series. — These  series  or  phenomena  described  by 
them  have  been  dealt  with  at  various  stages  of  our  argument.  We 
return  to  the  subject  merely  in  order  to  round  it  off,  particularly  with 
reference  to  banking  and  bankers'  banks'  policy.  For  this  purpose,  it  is 
convenient  to  define  the  stock  exchange  as  that  market  which  deals  in 
bonds  and  shares.  Another  part  of  the  market  of  "capital  values,"  the 
realty  market,  will  not  be  considered  here  in  spite  of  its  considerable 
cyclical  importance. 

Although  the  stock  exchange  is,  as  has  been  stated  in  Chap.  XII, 
really  a  part  of  what  in  the  widest  sense  we  have  called  the  open  market, 
it  will  now  be  looked  upon  as  distinct  from  it,  though  communicating 
with  it.  We  know  where  the  money  comes  from  that  buys  the  various 
"branded  articles"  that  constitute  the  "commodities"  of  this  very 
imperfect  market.  It  comes,  first,  from  the  "surplus  funds"  of  banks, 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    679 

lending  to  the  stock  exchange  being  the  most  important  method  of  banks' 
temporary  investment.  Second,  it  comes  from  nonbanks  (although  often 
through  bank  loans  on  account  of  others1),  domestic  business  concerns, 
and  foreign  banks  and  capitalists,  who  also  want  to  invest  funds  tem- 
porarily— the  latter  source  sometimes  enabling  member  banks  to  thwart 
bankers*  banks'  policies,  and  the  stock  exchange  to  thwart  member 
banks.  Third,  it  comes  from  "investors" — from  the  savings  of  house- 
holds and  from  the  funds  of  nonbank  firms  and  of  banks  which  also  may 
wish  to  invest  permanently.  Finally,  banks  directly  finance  trans- 
actions of  both  speculation  and  investment  by  collateral  loans  to  cus- 
tomers, which,  however,  must  not  in  general  be  identified  with  financing 
speculation  or  households'  investments  although  we  now  shall  tentatively 
assume  that  this  is  admissible  in  the  case  of  the  collateral  loans  of  New 
York  banks.  Though  all  banks  lend  for  stock  exchange  purposes,  their 
relation  to  and  position  in  the  stock  exchange  varies,  as  we  have  seen, 
very  much  in  time  and  as  between  countries,  and  this  difference  does  not 
disappear  if  we  include  among  banking  firms,  as  we  really  ought  to, 
brokers  and  jobbers  or  Paris  agents  de  change.  They  still  essentially 
remain  middlemen  and  never  have  the  position  of  German  banks.  The 
difference  between  a  speculator  and  an  investor  can  be  defined  by  the 
presence  or  absence  of  the  intention  to  "trade,"  i.e.,  to  realize  profits 
from  the  fluctuations  in  security  prices.  But  since  investors  also  bor- 
row and  since  they  may  at  any  time  turn  into  speculators,  this  does  not 
overcome  the  difficulty  of  distinguishing  speculative  from  nonspeculative 
transactions.  More  useful  for  practical  purposes  is  the  criterion  of  the 
margin  account. 

1.  The  shifting  of  old  issues  between  speculators,  investors,  and 
banks,  for  a  variety  of  purposes  with  which  we  are  familiar  and  which 
form  an  essential  link  in  the  mechanism  of  credit  creation,  constitutes 
one  group,  the  placing  of  new  issues  with  speculators  and  investors, 
another  group  of  transactions  relevant  for  our  purposes.  Disregarding, 
for  the  moment,  the  latter  group,  we  will  first  consider  the  role  in  the 
cycle  of  transactions  in  "old"  stocks  and  bonds  and  of  the  financing  of 
these  transactions.  Needless  to  say,  the  phenomenon  would,  in  the 
absence  of  external  disturbances,  owe  its  existence  entirely  to  the  cycle, 
and  must  obviously  be  expected  to  be  one  of  the  most  regular  of  its  fea- 
tures, so  much  so  that  it  would  be  hard  to  find  instances  of  cycles  that 
would  not  display  it.  It  is  clear,  moreover,  that  the  dissaving  implied  in 
living  on  gains  from  speculative  or  nonspeculative  transactions  of  that 
kind2  has  much  to  do  with  the  amount  of  consumers'  spending  in,  and 

1  The  role  of  Temporary  Investment  and,  in  particular,  of  Brokers'  Loans  will  be  dealt 
with  in  Chap.  XIV,  Sec.  F. 

2  The  gains  of  speculators  include,  no  doubt,  an  operational  element  which  is  not  a  true 
capital  gain  but  of  much  the  same  nature  as  any  other  return  to  personal  activity — for 


680  BUSINESS  CYCLES 

with  the  psychic  atmosphere  of,  prosperity,  and  that,  similarly,  losses  in 
recession  or  depression  have  much  to  do  with  the  reduction  of  consumers' 
spending  in,  and  with  the  psychic  atmosphere  of,  recession  or  depression. 
Nor  need  we  stay  to  show  how  increases  and  decreases  in  the  value  of 
collateral  may  be  responsible  for  loans  (also  for  business  purposes)  which 
would  not  have  been  granted  or  would  not  have  become  endangered 
without  them.  As  everybody  knows,  this  may  be  an  important  element 
in  the  banking  situation  of  "crises*5  and  in  vicious  spirals  in  general. 

Peculiarities  of  the  pricing  process  on  the  stock  exchange  and  the 
effects  of  speculation  on  the  situation  of  member  and  bankers'  banks  and 
on  the  financing  of  business  are  conveniently  dealt  with  by  reference  to 
the  time-honored  question  as  to  whether  "the  stock  exchange  absorbs 
credit."1  The  naive  argument  in  favor  of  an  affirmative  answer  has  been 
met,  and  for  practical  purposes  can  be  met,  by  pointing  out  that,  norm- 
ally though  not  in  times  of  speculative  manias,  "funds"  lent  on  the  stock 
exchange  would,  to  a  great  extent,  be  otherwise  unemployed.  Our 
previous  analysis  of  the  behavior  of  banks  in  cycles  lends  some  support  to 
this.  The  fact  that  in  the  presence  of  high  renewal  rates,  banks  and  other 
lenders  turn  from  the  commercial-paper  market  to  the  stock  market  does 
not  substantially  invalidate  that  reply,  nor  does  the  further  fact — the 
importance  of  which  greatly  varies  according  to  the  technique  of  specula- 
tion in  different  countries — that  anyone  who  wants  to  speculate  must 
"buy  himself  in,"  which  may  imply  withdrawal  of  existing  balances  or 
facilities  of  creating  them  from  other  uses,  without  necessarily  releasing  an 
equal  amount  at  once. 

More  fundamental  is  it,  however,  that  rising  stock  prices  are  not  in 
the  same  way  associated  with  increasing  use  of  balances  as  are  rising 
prices  of  ordinary  commodities.  Obviation  plays  a  greater  role2  than 
it  can  play  in  the  latter  case,  in  New  York,  particularly,  since  the  founda- 
tion in  1919  of  the  Stock  Clearing  Corporation.  Moreover,  brokers' 
loans  induce  no  cashing  of  checks  and,  consequently,  no  loss  of  cash  to 
circulation.  Again,  most  of  that  portion  of  brokers'  deposits  that  is 
derived  from  l&ans  need  not  actually  appear  as  deposits  at  all,  because  it 
is  entered  directly  under  the  banks'  certified  checking  accounts,  against 

instance,  the  professional  income  of  a  lawyer.  Spending  this  on  consumers'  goods  is  not 
dissaving.  There  is  no  need,  however,  to  emphasize  that  element  here. 

1  There  is  a  large  literature  on  this  question.     We  will  only  mention  Professor  Mach- 
lup's  Borsenkredit,  Industriekredit  und  Kapitalbildung,  1931;  R.  N.  Owens  and  C.  O. 
Hardy,  Interest  Rates  and  Stock  Speculation,  1925;  Thomas  Balogh,  Absorption  of  Credit 
by  the  Stock  Exchange,  American  Economic  Review  for  December  1930;  an  interesting  dis- 
cussion by  Professors  Cassel,  Spiethoff,  and  Hahn  in  the  Frankfurter  Zeitung  in  May  1927; 
and  R.  Reisch,  Riickwirkungen  der  Borsenspekulation  auf  den  Kreditmarkt,  Zeitschrift  fur 
Nationalokonomie,  1929.     The  question  will  be  taken  up  again  in  the  next  chapter,  sec.  F, 

2  See  Professor  Machlup,  op.  cit.>  p.  79,  et  seq. 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    681 

which  no  reserve  is  required.1     These  statements  may,   with  proper 
modifications,  be  paralleled  for  European  stock  exchanges. 

But  the  essential  point,  which  really  arises  out  of  the  same  circum- 
stances and  also  helps  to  account  for  obviation's  being  so  easy  in  this  case, 
is  this.  What  we  have  termed  the  efficiency  of  money  bears  a  rela- 
tion, fundamental  to  any  theory  of  money,  to  institutional  periods  of 
payment,  which,  in  turn,  bear  a  relation  to  the  timing  and  periodicities  of 
economic  processes.  There  is  no  such  time  element  in  transactions  on  the 
stock  exchange,  because  there  is  no  process  to  go  through.  Institutional 
arrangements  and  the  limits  of  physical  possibility  still  produce  a  mini- 
mum period,  but  it  might  conceivably  be  one  of  minutes  only.  Hence, 
there  is  almost  no  limit  to  a  "price  level"  that  a  given  "fund/*  however 
small,  may  "support."  Moreover,  while  a  commodity  must,  at  the 
price  determined  in  its  market,  move  over  its  track  through  the  economic 
organism,  there  is  no  such  necessity  for  stocks.  If,  in  the  case  of  a 
commodity,  market  price  is  such  as  to  prevent  it  from  starting  on,  or  com- 
pleting, its  course,  stocks  of  the  commodity  accumulate,  which  in  them- 
selves are  irksome  maladjustments  and  must  eventually  be  liquidated. 
This  keeps  each  price  ultimately  in  line  with  other  prices  and  in  subjection 
to  the  monetary  ligarnen.  But  in  the  case  of,  say,  a  common  stock,  the 
fact  that  at  its  price  it  cannot  "move"  docs  not  necessarily  spell  an 
untenable  situation.  The  people  whose  action  and  estimation  is  responsi- 
ble for  that  price  may  be  willing  to  keep  it,  and  normally  can  do  so. 
The  stock  need  not  move  to  any  definite  spot  in  the  economic  organism,  in 
order  to  fill  its  economic  function.  And  the  mere  fact  that  a  small  group 
of  people  entertain  an  irrationally  high  opinion  about  the  value  of  a  stock 
can,  with  very  small  transactions,  raise  it  to  indefinite  heights  without 
either  absorbing  any  funds  or  credit  facilities  or  eliciting  pressure  from 
monetary  magnitudes.  This  is  the  great  difference  between  pricing  on 
stock  and  pricing  on  produce  exchanges.2  Technical  qualifications  apart, 
into  which  we  need  not  enter,  such  as  the  holding  of  war  chests  by  pools 
and  individuals,  it  is,  therefore,  true  that  stock  speculation  docs  not 
absorb  money  or  capital  or  credit  in  the  sense  usually  implied  by  this 
phrase.3  New  issues,  of  course,  do  "absorb  funds,"  but  only  to  release 
them,  unless  the  proceeds  be  applied  to  the  repayment  of  bank  loans. 

1  H.  Rogers,  Effect  of  Stock  Speculation  on  the  New  York  Money  Market,  Quarterly 
Journal  of  Economics,  vol.  XL,  p.  453. 

2  It  has  been  pointed  out  that  a  similar  peculiarity  exists  in  many  cases  of  commodities 
in  the  proper  sense  of  the  word,  such  as  pictures  by  old  masters  and  real  estate.     This  is 
quite  true,  even  though  certain  other  characteristics  of  the  stock  market  are  absent.     But 
this  fact  does  not  constitute  an  objection — it  rather  illustrates  our  meaning. 

3  This  view,  together  with  some  of  the  propositions  that  are  to  follow,  is  often  taken  to 
imply  "friendliness"  to,  or  an  intention  to  "defend,"  stock  speculation.     In  order  to  dispel 
any  such  suspicion,  the  writer  begs  leave  to  state  that  on  moral,  political  and  cultural 


BUSINESS  CYCLES 

2.  In  analyzing  the  pricing  process  on  the  stock  exchanges,  another 
consequence  of  the  fact  that  the  "commodities"  of  this  market  are,  by 
virtue  of  their  economic  nature,  held  rather  than  moved  must  be  taken 
into  account.  A  market  in  which  this  is  the  case  and  in  which  the  pres- 
sure of  the  monetary  ligamen  is  not  present  in  the  same  sense  as  it  is  in 
markets  of  ordinary  commodities,  in  which,  moreover,  stocks  are  to  a 
great  extent  held  for  a  rise  but,  serving  as  collateral,  have  to  be  liquidated 
in  case  of  a  fall  in  price,  in  which  "  supply "  will  normally^  increase  in 
consequence  of  a  fall,  decrease  in  consequence  of  a  rise,  or  also  vary 
without  any  variation  in  price — such  a  market  is  not  likely  to  conform  to 
traditional  ideas  of  the  behavior  of  supply  and  demand.  If  it  does  not 
behave  more  erratically  than  it  does,  this  is  due  to  the  fact,  easy  to  verify, 
that  the  varied  responses  which  a  change  in  the  situation  elicits  in  part 
balance  each  other.  The  resulting  tone  or  tendency  is,  as  a  rule,  so 
clearly  marked,  because  it  is,  as  soon  as  discerned,  acted  upon  by  specula- 
tors. In  a  sense  this  is  true  of  any  market;  but  for  the  reasons  stated 
above,  it  holds  for  no  other  market  to  the  same  extent. 

Therefore,  explanation  of  that  tone  or  tendency  has  been  rightly  felt 
to  be  a  distinct  task,  i.e.9  a  task  not  only  distinct  from  the  explanation  of 
market  phenomena  in  general  but  also  of  the  phenomena  in  other  specu- 
lative markets.  Speculation  in  raw  material  futures  is,  for  instance, 
much  more  likely  than  speculation  in  common  stock  to  display  those 
features  that  can  be  listed  under  the  traditional  and  self-explanatory 
headings;  insurance,  arbitrage,  and  steadying  price  over  time.  Carrying 
temporarily  stock  that  people  do  not,  or  do  not  yet,  care  to  carry  perma- 
nently, and  sending  monetary  capital  into  the  firing  line  come  nearer  to  a 
description  of  what  speculation  in  stocks  actually  does.1  What  follows 
refers  primarily  to  the  behavior  of  stock  prices.  The  behavior  of  bond 
prices  has  been  dealt  with  in  the  preceding  chapter.  But  it  must  be 
remembered  that  the  pricing  of  low-grade  bonds  involves  a  speculative 
element,  which  will  tend  to  make  their  prices  behave  much  like  prices  of 
stocks. 

Since  stock  prices  have  more  degrees  of  freedom  than  other  prices 
have,  and  since  financial  groups — pools  and  others — confront  a  public 
very  much  more  excitable  and  very  much  less  intelligent  than  the  constit- 
uent individuals  are  in  their  ordinary  business  pursuits,  it  is  tempting  to 
stress  mere  mass  psychology,  on  the  one  hand,  and  mere  abundance  or 

grounds  he  personally  welcomes  almost  any  measure  in  any  way  hostile  to  it,  regardless  of 
economic  consequences.  In  particular,  he  would  welcome  an  enactment  making  specula- 
tion a  misdemeanor  for  the  members  of  certain  professions. 

1  The  theory  that  no  concern  is  a  permanent  source  of  net  revenue  allocates,  however, 
as  will  be  clear  on  reflection,  to  trading  in  stocks  a  significance  within  the  capitalist  process 
which  is  additional  to  the  "functions"  of  it  that  enter  into  traditional  formulations. 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    683 

scarcity  of  funds,  on  the  other.  While  there  cannot  be  any  doubt  about 
the  presence  of  either  element,  neither  of  them  makes,  by  itself  or  jointly 
with  the  other,  a  satisfactory  explanation.  As  regards  the  first,  there  is, 
of  course,  much  more  scope  for  waves  of  optimism  and  pessimism  on  the 
stock  exchange  than  there  is  in  industrial  and  commercial  business,  and 
it  is  more  nearly  true  for  the  former  than  it  is  for  the  latter  that  anticipa- 
tions may  temporarily  produce  the  expected  result.  But  even  manias 
have  their  starters,  and  we  must  not  judge  by  manias  only.  It  should  be 
added  that  there  has  been  a  slow  but  persistent  change  in  the  behavior  of 
the  public,  which  becomes  obvious  if  we  distinguish  irrational  optimism 
as  to  general  conditions  and  irrationality  in  the  choice  of  the  particular 
stock  to  speculate  in.  The  scene  of  1929,  perhaps,  differed  but  little 
from  the  scene  of  18731  or  even  from  that  of  the  South  Sea  Bubble,  in 
the  former  respect;  but  it  differed  significantly  in  the  latter.  In  order 
to  choose  rationally,  a  man  need  not  be  able  to  analyze  a  business  situ- 
ation or  the  state  and  prospects  of  an  industry  or  concern.  All  he  needs 
to  know  is  what  advice  or  example  or  current  opinion  to  follow.  It 
remains  true  that  irrational  fancy  and  downright  foolish  hopes  or  fears 
count  for  much  in  the  short  run.  But  it  is  no  less  true  that  they  never 
provide  the  motive  power  of  a  boom  and  that  they  never  prevent  the 
real  state  of  things  from  asserting  itself  eventually.2  This  real,  or  objec- 
tive, state  may  be  decomposed  into  external  factors,  cyclical  phases,  and 
conditions  peculiar  to  individual  industries  and  concerns.  No  doubt, 
the  word  expected  may  usefully  be  inserted  in  each  case. 

While  the  above  analysis  of  the  peculiarities  of  stock  pricing  thus 
gives  some,  if  limited,  support  to  an  optimism -pessimism  theory,  it 
rather  suggests  that  less  emphasis  should  be  placed  on  money-market 
factors  than  we  should  infer  from  the  familiar  statistical  evidence.  We 
have  seen  in  the  preceding  chapter  that  small  variations  in  open-market 
rates  made  a  great  deal  of  difference  to  the  operator  in  bonds  and  hence 
to  their  prices,  but  no  variations  of  rates  such  as  currently  occur  in  a 
normal  cycle  can  possibly  matter  for  speculation  in  stock.  It  is,  there- 
fore, availability  rather  than  cost  of  credit  that  we  should  look  to.  But 
we  have  seen  that,  barring  new  issues,  comparatively  small  amounts  will 
go  a  long  way  on  the  stock  exchange.  Considerable  booms  can  develop 

1  Even  that  is  not  quite  certain,  however.  Foreign  buyers  in  1929  need  not  have  been 
unduly  optimistic  as  to  American  conditions  in  order  to  have  preferred  American  stocks  to 
European  ones.  And  if  American  holders  of  American  stocks  failed  to  sell  in  time,  this 
does  not  prove  that  they  foresaw  nothing  but  continuing  booms,  but  only  that  they  did  not 
foresee  that  the  loss  from  a  fall  in  security  prices  would  be  greater  than  the  amount  of 
income  tax  which  would  have  been  due  had  they  sold  out. 

8  The  stock  exchange  thus  affords  an  excellent  example  by  which  to  demonstrate  what 
"psychological  factors"  and,  in  particular,  "expectations"  can  and  cannot  do  and  in  what 
sense  economic  reality  is  independent  of  them. 


684  BUSINESS  CYCLES 

on  a  narrow  basis  of  "cash,"  and  unless  banks  set  their  faces  with  what 
would  be  quite  unusual  determination  against  the  speculative  purpose,  it 
is  not  easy  to  see  how  speculation  could  be  starved  by  lack  of  funds, 
though,  of  course,  new  issues  could.  Hence,  we  shall  be  inclined  to 
attribute  that  statistical  regularity  more  to  coincidence  within  the  cyclical 
process — to  be  explained  presently — than  to  the  causative  influence  of 
monetary  factors,  for  example,  in  the  sense  of  the  theory  according  to 
which  funds  migrate  from  industry  and  commerce  into  the  stock  exchange, 
thereby  igniting  speculative  booms,  and  then  back  again,  thereby  putting 
a  stop  to  then. 

3.  Indices  of  stock  prices  make  synthetic  and  systematic  series 
descriptive  of  a  primary  and  consequential  phenomenon.  These  series 
will,  of  course,  be  strongly  cyclical,  however  they  may  be  constructed. 
But  our  model  yields  no  generalization  about  trend.  Such  trends  as  we  find 
simply  reflect  historical  facts,  which  must  be  separately  accounted  for  in 
each  case,  and  depend  on  the  properties  of  the  index  chosen.  If,  for 
example,  the  index  expressed  average  price  per  dollar  of  paid-up  capital 
plus  accumulation  for  a  number  of  identical  concerns,  it  would,  barring 
changes  in  the  purchasing  power  of  the  monetary  unit  and  in  the  value  of 
natural  agents  owned  by  those  concerns,  normally  go  to  zero  after  a 
period  of  sufficient  duration.  If  new  concerns  are  currently  added  and 
old  ones  currently  dropped,  it  will  not  do  that,  but  still  need  not  display 
any  particular  result  trend  in  our  sense. 

Since  we  do  not  believe  in  the  dominant  role  of  money-market  condi- 
tions we  shall  not  expect  any  strong  seasonal  variation,  although  holidays 
and  so  on  must  exert  some  influence.  As  regards  cyclical  variation,  three 
facts  must  be  borne  in  mind.  First,  the  public's  anticipations  concerning 
cyclical  phase  and  its  anticipations  concerning  the  fortunes  of  individual 
industries  and  concerns  are  interdependent  and  tend,  within  the  short  or 
medium  run,  relevant  to  the  decisions  of  speculators  and  even  investors, 
to  point  in  the  same  direction.  Even  an  old  and  ailing  concern  will  in 
the  majority  of  cases  do  better  in  prosperity  than  in  depression.  But 
both  speculators  and  investors  are  particularly  attracted  by  relatively 
new — though  perhaps  not  so  much  by  quite  new  and  untried — industries 
and  by  the  leaders  in  those  industries  which  carry  prosperities.  That  this 
is  so  is  clear — railroad,  electricity,  motor,  rubber,  oil  stocks  and  dozens 
of  equally  familiar  instances  lie  at  hand  to  verify  it.  But  it  should  be 
noticed  how  well  this  fact  illustrates  the  "function"  and  also  the  "ration- 
ality" of  speculation;  for  the  diagnoses  which  this  behavior  implies  are  in 
general  well  verified  by  objective  events,  themselves  independent  of  those 
anticipations.  Not,  of  course,  in  all  individual  cases,  nor  necessarily  from 
the  standpoint  of  the  speculator — but  even  when  the  latter  loses  his 
money,  he  rarely,  if  ever,  acts  on  dreams:  the  developments  he  counts 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE     685 

upon  almost  invariably  mature,  however  seriously  he  may  be  at  fault  as 
to  timing. 

Second,  the  stock  market  is  imperfect,  but  it  is  freer  from  friction 
than  almost  any  other.  Decision  can  be  given  effect  to  and  reversed 
with  a  promptness  altogether  impossible  in  industry.  Depressive 
situations  may  so  weaken  hands  as  to  force  them  to  relax  their  grip,  even 
after  opinions  have  undergone  a  favorable  change,  but  on  the  other  hand, 
the  effects  of  such  a  change  may  be  accentuated  by  the  supply  being 
suddenly  curtailed  when  demand  revives.  Hence,  it  is  natural  to  expect 
that  upward  movements  on  the  stock  exchange  will,  in  general  and  in  the 
absence  of  unfavorable  external  factors,  set  in  earlier  and  gather  force 
more  quickly  than  the  corresponding  upward  movements  in  business, 
i.e.,  often  come  about  already  in  the  later  stages  of  revival  when  things  are 
beginning  to  look  better  every  day,  with  new  possibilities  showing  them- 
selves. Similarly,  it  is  to  be  expected  that  stock  prices  will  turn  before 
other  indicators,  i.e.,  when  in  the  later  stages  of  prosperity  limitations 
and  difficulties  emerge  and  it  becomes  clear  that  possible  achievements 
have  been  fully  discounted.1  Now,  in  the  first  situation,  interest  is  low 
and  the  money  and  open  markets  are  normally  easy;  in  the  second, 
interest  is  high  and  the  money  and  open  markets  are  normally  tight. 
Some  causal  significance  may  be  attributed  to  this  relation,  particularly 
if  we  include  the  role  that  monetary  conditions  will  play  in  the  prognosis 
by  speculators  or  their  advisers  of  what  is  ahead  in  business.2  But  the 
principal  explanation  is  that  both  stock  prices  and  money  rates  respond 
to  cyclical  phases  so  strongly  that  their  variations  would  be  regularly 
associated  even  if  they  had  nothing  to  do  with  each  other. 

Third,  because  of  that  promptness  in  reaction  and  also  because  the 
self-reenforcing  mechanism  is  so  strong  in  the  stock  market,  we  must 
expect  movements  to  outrun  their  momentary  goals  much  more  than  in 
other  markets  and  reactions  to  be  correspondingly  sharp.  Even  if  the 
market  be  undecided,  this  will  mean  fluctuations — of  the  sort  we  called 
hesitations  in  Chap.  IV,  Sec.  E — rather  than  inaction.  Moreover,  while 
general  business  may  and  often  does  settle  down  into  recession  in  a  per- 
fectly orderly  way,  this  is  hardly  imaginable  in  the  case  of  the  stock 
exchange.  Recession  means  reduced  profits  and,  for  many  concerns, 
more  or  less  serious  troubles.  It  gives  scope  to  bear  attacks.  But  even 

1  That  view  is  supported  by  the  observation  that  very  frequently  the  rise  in  the  leading 
stocks  stops  before  the  top  of  the  market  is  reached.     Speculative  interest,  realizing  that 
their  possibilities  have  for  the  time  being  been  exhausted,  looks  around  for  stragglers  that 
may  still  offer  secondary  chances.     That  is  when  the  "froth"  comes  up. 

2  The  writer  is  inclined  to  believe  that  the  facts  that  the  weekly  statements  of  the  New 
York  Associated  Banks  used  to  be  a  feature  in  the  Saturday  market  and  that  similar  atten- 
tion was  in  London  and  Berlin  paid  to  the  statements  of  the  Bank  of  England  and  of  the 
Reichsbank,  were  entirely  due  to  this. 


680  BUSINESS  CYCLES 

if  nothing  of  this  sort  ever  happened  or  were  anticipated,  the  mere  fact 
that  there  is  no  reason  to  expect,  save  in  particular  cases,  any  upward 
movement  would  suffice  to  make  speculators  lose  interest  in  their  hold- 
ings. Thus  the  upper  turning  point  or  the  approach  of  it  is,  in  this  case, 
likely  to  mean  a  crash,  which  need  not  greatly  affect  the  general  business 
situation.  But  although  it  need  not,  it  often  will,  both  by  upsetting  the 
credit  situation  and  by  uncovering  weak  spots.  It  is  easy  to  carry  this 
argument  through  spirals  and  depressions  into  recovery  and  to  satisfy 
oneself  that  in  the  absence  of  an  abnormal  amount  of  reckless  finance  and 
misconduct  and  often  in  spite  of  their  presence,  bear  markets  do  not  last 
unless  coinciding  with  and  independently  induced  by,  depressions  in 
general  business.  This  is  perfectly  compatible  with  the  influence,  pre- 
viously recognized,  which  stock  exchange  events  exert  on  consumers* 
spending. 

For  corroboration  we  return  to  Chart  XXXVI.  The  fairly  marked 
covariation  of  railroad  stock  prices — as  long  as  these  were  the  dominant 
element,  i.e.,  to  the  nineties — and  of  railroad  and  industrial  stock  prices 
with  New  York  loans  and  deposits,  and  their  strongly  marked  covariation 
with  New  York  Clearings1  should  again  be  noticed  first.  In  both  railroad 
and  industrial  stock  prices  the  Kondratieff  prosperity  from  1897  on 
shows  well  and  so  do  the  Kitchins.  The  major  movements  which  we 
observe,  however,  clearly  reflect  the  Juglars:  we  see  the  (anticipating) 
boom  of  1868  and  1869  and  then  the  characteristic  slump  from  1873  to 
1877;  then  the  (also  anticipating)  boom  from  1877  to  1881;  the  same 
phenomenon,  most  regularly  repeated  from  1885  on;  no  such  precedence 
for  the  first  Juglar  of  the  third  Kondratieff,  which  may  have  been  due  to 
the  aftereffects  of  1893  and  political  factors;  but  more  regular  behavior 
again  in  the  second.  This  reflects  the  relation,  which  according  to  our 
analysis  should  exist  between  speculation  and  the  investment  process. 
The  latter,  as  we  know,  dominates  the  Juglars  more  obviously  than  the 
Kitchins.  All  series  that  link  up  with  investment,  such  as  pig-iron 
production,  employment,  and  so  on,  display  Juglars  with  particular 
emphasis.  The  fact  that  stock  speculation  behaves  similarly  is,  hence, 
of  some  significance.  The  reader  should  again  inspect  Chart  XXXIV, 
particularly  the  graph  of  New  Listings,  for  which  speculation  paves  the 
way. 

Charts  XXXVII  and  XXXVIII  substantially  tell  the  same  tale.  In 
the  first  there  should  be  noted  the  instructive  differences  that  comparison 
with  the  American  series  reveals,  among  them  the  spurt  in  stock  prices 
from  1895  to  1897,  strongly  underlined  by  the  behavior  of  call  rate,  after 
which  they  fluctuated  around  a  declining  level.  The  second  presents  a 

1  See  A.  M.  Matthews,  New  York  Bank  Clearings  and  Stock  Prices,  1866-1914,  Review 
of  Economic  Statistics  for  October  1926,  where  trend  ratios  are  compared. 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    687 


picture  which  is  much  more,  though  not  quite,  similar  to  the  American. 
German  stock  prices  reproduce  the  movement  of  American  industrials  up 
to,  and  down  from,  the  peak  of  1872.  In  reality,  however,  this  veils  a 
difference  in  behavior,  since  the  German  index  includes  all  leaders,  while 
the  American  index  excludes  railroad  stocks,  which  in  this  country,  more 
agreeably  to  expectation  reached  their  peak  sooner,  i.e.,  in  1869.  Political 
events — the  Franco-Prussian  War — may  plausibly  account  for  this  differ- 
ence. The  trough  occurs  in  the  same  year  as  in  this  country,  1877, 
but  the  subsequent  rise  already  stopped  in  1880,  the  rise  of  1888  in 


1870 


1875 


I860  1885  1890  1895  1900 

CHART  XXXVII— (See  Appendix,  p.  1064.) 


1905 


1910 


1914 


1889,  after  which  the  index  falls,  preceding  the  American  index  to  the 
end  of  1892.  Our  guess  that  political  events  interfered  with  the  course 
of  things  after  1893,  this  time  in  America,  is  supported  by  the  fact  that 
in  Germany  a  strong  upward  movement,  which  lasted  through  1898,  had 
already  begun  in  1894.  This  would  make  a  very  normal  case,  not  at  all 
impaired  by  the  duration  of  the  interval — over  three  years — by  which  the 
stock  exchange  shadow  preceded  the  form  that  cast  it,  viz.,  the  prosperity 
which  we  date  from  1898.  We  need  only  look  at  the  general  situation, 
which  may  be  described  as  buoyant  revival,  and  at  the  stock  exchange 
leaders:  everything  was  in  fact  "looking  better  every  day"  and  the  con- 
tours of  the  achievements  in  the  electrical  field  unmistakably  stood  up 
before  everyone's  eyes.  By  the  end  of  1898,  on  the  other  hand,  it 
was  not  less  clear  that  possibilities  were  for  the  moment  being  rapidly 
exhausted — more  than  exhausted — and  that  the  immediate  future  offered 
little  for  the  fancy  of  speculation  to  feed  on. 


688 


BUSINESS  CYCLES 


THE  CENTRAL  MARKET  AND  THE  STOCK  EXCHANGE    689 

No  undue  amount  of  conscious  rationality  is  by  this  interpretation 
attributed  to  the  rank  and  file  of  speculators.  But  it  is  held  that  objec- 
tively there  was  method  in  the  madness.  In  this  respect  it  is  also  worth 
while  observing  how  close  stock  prices  come  to  reproducing  the  move- 
ment of  dividends.1  And  both  go  fairly  well  with  Foundations  and 
Issues  of  Industrial  Stock.  While  it  is  unnecessary  to  dwell  on  the 
significance  of  this  fact,  a  few  minor  points  should  be  noticed.  Founda- 
tions and  Issues  do  not  display  as  close  a  covariation  as  one  might  expect; 
on  reflection,  however,  it  will  be  realized  that  the  former,  being  of  the 
essence  of  industrial  booms,  will  cluster  at  or  before  the  peaks  of  business 
activity,  while  the  latter,  being  tied  to  the  opportunities  created  by 
speculation,  will  cluster  at  or  before  the  peaks  of  stock  exchange  activity. 
Issues  of  industrial  bonds  and  issues  of  stocks  should  display  similar 
" trends"  but  opposite  fluctuations.  The  former  expectation  is  verified, 
but  the  relation  between  the  fluctuations  is  less  consistent  than  we  might 
have  hoped.  A  peak  occurs  in  both  series  in  1911.  Juglars  show  excel- 
lently in  stock  issues,  Kitchins  rather  better  in  bond  issues. 

4.  In  this  and  in  earlier  chapters  facts  and  arguments  have  been  pre- 
sented which  suffice  to  draw  the  contours  of  a  theory  of  the  relation  of 
banks  to  stock  speculation  and  to  that  part  of  the  investment  process  of 
which  the  stock  exchange  is  the  scene.  We  may  sum  up  by  stating  that 
even  where  banks  most  strictly  keep  to  the  principles  of  classic  "deposit 
banking,"  considerations  arising  out  of  the  practice  of  lending,  to  the 
stock  exchange  and  to  customers,  on  stock  exchange  securities  provide  a 
most  important  link  between  the  prices  of  the  latter  and  the  fortunes  of 
each  individual  bank;  but  that  where  banking  of  the  credit  mobilier  type 
prevails  and  banks  directly  "patronize"  industries,  control  their  financial 
policy,  manage  their  issues,  hold  stocks  or  bonds  and  trade  in  them  on 
their  own  account,  the  stock  exchange  may  become  the  pivot  of  their 
business.  But  we  will  now  confine  ourselves  to  the  relation  of  bankers' 
banks  to  the  stock  exchange. 

From  the  proposition  that  speculation  in  stocks  does  not,  or  not  to 
a  significant  extent,  "absorb  credit" — that  the  stock  exchange  is  not  a 
sponge  but  a  channel — it  does  not  follow  that  it  is  a  matter  of  indifference 
to  the  central  bank.  It  is  obvious  that  new  issues  are  not.  For  they 
precisely  supply  what  it  is  the  function  of  a  central  bank  to  regulate. 
They  may  increase  the  lending  power  not  only  of  individual  member 
banks  but  of  the  banking  system  as  a  whole.  As  we  have  previously 
seen,  they  may,  if  for  foreign  account  or  merely  if  internationally  financed, 

1  Prices  of  the  stocks  of  individual  concerns  are  also — everywhere — closely  associated, 
though  only  in  the  short  run,  with  the  prices  of  the  products  of  those  concerns.  It  would 
be  dangerous  to  stress  this  relation  too  strongly.  It  points,  nevertheless,  to  the  influence 
of  obvious  common-sense  considerations  on  the  decisions  of  speculators.  The  relation 
is  instantaneous,  however,  and  this  type  of  decision  is  likely  to  be  taken  postfestum, 


690  BUSINESS  CYCLES 

upset  the  working  of  the  mechanism  of  short  balances  and  affect  gold 
movements  and  the  central  banks'  reserve.  Barring  this,  they  always 
affect  the  open-market  situation.  And  they  are  associated  with  reckless 
financing.  It  stands  to  reason  that  the  central  bank  cannot  be  indifferent 
to  that. 

But  if  central  banks  cannot  be  indifferent  to  the  volume  and  the  pur- 
poses of  new  issues,  they  cannot  be  indifferent  to  the  movement  of  stock 
prices,  because  a  boom  in  these  is,  practically  speaking,  the  necessary 
and  (almost)  the  sufficient  condition  for  a  boom  in  issues.  Moreover, 
even  trading  in  old  stocks  influences  business  activity — the  intensity  of 
the  Secondary  Wave — through  the  withdrawal  and  the  spending  of 
speculative  gains,  and  also  the  general  banking  situation,  especially  if 
member  banks  themselves  are  sellers  or  buyers  on  their  own  account. 
This  need  not  cross  and  may  even  help  the  policy  of  the  bankers'  bank, 
as  we  have  seen.  We  have,  in  particular,  had  occasion  to  observe  that 
the  selling  by  banks  of  stocks  and  bonds  may  effectively  restrain  a  boom 
and  that  their  buying  may  mitigate  a  slump.  But  this  need  not  be  so. 
It  follows  that  the  untenable  theory  about  speculation  absorbing  funds 
may  be,  after  all,  the  ill-tailored  coat  of  much  practical  wisdom :  though 
for  a  different  and,  in  fact,  almost  opposite  reason,  there  are  plenty  of 
motives  for  bankers'  banks  to  react  to  stock  speculation  and  to  try  to 
fight  its  excesses  during  prosperity  or  the  later  stages  of  recovery,  and 
vice  versa  bear  attacks  during  recession  and  depression.  Success  in  this 
may  avoid  many  hitches  in  the  investment  process,  reduce  many  plus  or 
minus  deviations  of  current  business,  remove  some  of  the  darkest  hues 
from  the  picture  of  "deep"  depressions. 

The  partial  analogy  that  exists  between  new  issues  and  open-market 
operations  by  the  bankers'  bank  suggests  that  the  former  may,  to  some 
extent  at  least,  be  controlled  by  the  latter  or,  more  generally,  be  regulated 
by  a  central  bank  that  keeps  a  firm  hold  on  the  open  market.  New  issues 
are,  in  fact,  the  most  important  link  between  stock  exchange  booms  and 
easy  money,  and  sensitive  to  tightness.  Moreover,  there  are  other  and 
more  direct  means  to  the  same  end  which  it  is  and  was  within  the  power 
of,  say,  the  Reichsbank  or  the  Bank  of  England  to  use.  But  acting  on 
the  conditions  that  facilitate  or  impede  new  issues  is  a  delicate  task.  It 
might  prevent  financial  consolidation  in  individual  points  of  potential 
danger  or  even  bar  the  way  to  consolidation  of  an  overwrought  situation 
in  the  banking  system  and  thus  precipitate  what  it  is  the  intention  to 
avoid.  In  fact  it  is  impossible  as  a  general  policy.  Some  effects  on  the 
economic  process  are,  moreover,  as  we  have  seen,  produced  by  booms  in 
the  prices  of  existing  stock,  independently  of  whether  or  not  they  induce 
new  issues.  It  has,  therefore,  been  at  all  times  the  almost  universal 
opinion  of  practitioners  of  central  banking  that  the  management  of 


THE  CENTRAL,  MARKET  AND  THE  STOCK  EXCHANGE    691 

credit  would  never  be  fully  effective  and  especially  never  effective  in  time, 
unless  stock  speculation  be  acted  on  directly. 

Nevertheless,  we  see  throughout  the  period  under  survey  very  little 
effort  to  accomplish  this  and  very  poor  results  of  such  efforts  as  were 
occasionally  made.  As  we  have  seen  in  Chap.  VII,  something  was  done  in 
Germany,  but  it  was  done  by  legislation  and  not  by  central  bank  action. 
The  reason  for  this  follows  from  our  analysis  of  pricing  in  the  stock  mar- 
ket. Since  stock  speculation  does  not  absorb  funds,  it  must  be  extremely 
difficult  to  stop  or  to  restrain  by  any  of  the  ordinary  tools  of  central  bank- 
ing. Bank  rate  in  particular  is  anything  but  omnipotent  with  respect  to 
production  and  commerce,  but  it  is,  within  the  limits  that  are  practically 
feasible,  almost  ineffective  with  respect  to  stock  speculation:  regulative 
influence  on  stock  prices  presents  a  problem  entirely  different  from  that  of 
influencing  commodity  prices. 


CHAPTER  XIV 

1919-1929 


A.  Postwar  Events  and  Postwar  Problems. — The  formidable  task  of 
interpreting,  economically  and  sociologically,  our  own  time  cannot  be 
attacked  in  this  book.  Still  less  is  it  within  our  plan  to  recommend  or  to 
criticize  remedial  policies  or  fundamental  reforms,  or  even  to  enter  into 
discussion  of  individual  measures  taken  or  proposed.  Whatever  of  this 
the  reader  may  find,  explicitly  or  implicitly,  in  the  following  pages,  is 
incidental  to  an  argument  the  very  restricted  purpose  of  which  should  be 
borne  in  mind  throughout.  That  purpose  is  to  answer  the  question  how 
far  the  cyclical  process  of  capitalist  evolution,  as  analyzed  for  the  130 
years  that  preceded  the  World  War,  can  be  proved  to  have  persisted  in 
the  postwar  period,  and  to  see  how  our  model  works  under  the  conditions 
and  with  the  richer  material  of  that  period.  The  contribution  toward 
an  understanding  of  the  postwar-world  which  an  investigation  of  this 
kind  can  be  expected  to  make  may  prove  worse  than  valueless,  if  its 
character,  methodological  background,  and  analytic  intention  are  allowed 
to  drop  out  of  sight.  Wherever  it  seemed  possible,  an  attempt  has  been 
made  to  save  space  and  to  rely  on  the  fact  that  current  economic  events 
are,  and  have  been  since  the  war,  very  much  more  efficiently  reported 
than  before,  and  on  the  hope  that  general  contour  lines  are,  therefore, 
familiar  to  the  reader. *  It  will  be  convenient  to  deal  first  with  the  years 
preceding  the  world  crisis  and  to  comment  on  the  latter  in  a  separate 
chapter. 

We  exclude  the  years  1914  to  1918— for  Germany  also  1918  to  1923 
— on  the  ground  that  those  years  were  dominated  by  "external  factors" 
to  an  extent  that  makes  their  figures  valueless  for  our  purpose.  This  is, 
indeed,  not  quite  true.  The  rhythm  of  economic  life  clearly  persisted  in 
the  United  States,  and  some  aspects  of  war  events  are  not  without  rele- 
1  It  is,  hence,  suggested  that  the  reader  refresh  his  memory  by  going  over  the  charts  and 
comments  of  one  of  the  well-known  services.  The  Harvard  Service's  quarterly  surveys 
are  particularly  recommended  (for  Germany,  the  surveys  of  the  Institut  fiir  Konjunktur- 
forschung,  for  England  those  of  the  London  and  Cambridge  Economic  Service).  For 
students  of  economics  there  is  no  more  stimulating  exercise  than  to  bring  to  bear  their 
analytic  apparatus  on  the  task  of  interpreting  the  course  of  postwar  events  as  described 
in  those  reports.  The  most  convenient  basis  for  a  study  on  a  world-wide  scale  is,  per- 
haps, the  material  of  the  League  of  Nations. 

602 


1919-1929 


693 


DEPOSITS  Pit 
CIRCULATION 


.US 


J 


/\  INTEREST  RATE  ON 
'•PRIME  COMMERCIAL  FftPER 


\  f\ 
\f  ' 

V     '. 


ft  a 


vance  for  the  study  of  business  cycles.  In  particular,  war  expenditure 
affords  as  good  experimental  evidence  as  we  can  ever  hope  to  get 
about  the  nature  and  consequences  of 
a  boom  which  has  nothing  to  do 
with  innovation  and  is  brought  about 
by  expanding  credit  and  stimulating 
consumption  alone.  The  fact  that 
expenditure  was  not  directed  into 
channels  which  would  commend  them- 
selves to  advocates  of  such  a  policy  is 
entirely  irrelevant,1  for  all  that 
matters  is  that  depressions  were  actu- 
ally impending  or  in  progress  in  1914 
and  that  public  expenditure  turned 
them  into  prosperity  first  and  created 
untenable  situations  afterward.  But 
although  the  case  almost  ideally  com- 
plements and  illustrates  part  of  the 
argument  of  this  book,  we  will  follow 
the  practice  of  the  majority  of  stud- 
ents and  eliminate  those  violent "  irreg- 
ularities" by  leaving  out  the  figures 
of  those  years. 

It  is  obvious,  however,  that  exter- 
nal factors  in  our  sense  continued  to 
play  a  supernormally  important  r61e 
throughout  the  postwar  period.  That 
our  second  component  of  economic 
change,  the  cyclical  process  of  evolu- 
tion, was  still  present  and  asserted 
itself  in  the  same  manner  as  before 
is  not  so  obvious.  Owing  to  the  his- 


I    I     I     I 


I     I     I 


I     I     I     I 


torical   character  of  our  subject — or  l919  2^  193°  3* 

the    fact    that    it    is    "institutionally  CHART  XXXIX.-United  States  postwar 

,.,.          ,,,      .,  .  ,.  ,  ,  '  pulse     (see  Appendix,  p.  1065). 

conditioned  — this     question    would 

arise  in  any  case,  even  if  there  had  been  no  war:  whenever  we  wish  to 
apply  our  analysis  to  an  additional  span  of  time,  we  must  always  ask 

1  Similarly,  it  is,  in  the  case  of  the  spending  policy  so  consistently  followed  by  the 
government  of  Louis  XV,  entirely  irrelevant  that  most  modern  advocates  of  "expansion" 
would  not  approve  of  the  objects  to  which  that  expenditure  was  applied.  Whatever  those 
objects,  such  expenditure  ought  to  have  produced  prosperity  without  relapse.  If  there  is 
anything  in  a  theory  which,  in  fact,  was  widely  held  at  that  time  and  has  experienced  a  most 
striking  revival  in  our  days,  it  would  be  but  justice  to  rewrite  the  history  of,  say,  the  minis- 
try Aiguillon-Maupeou-Terray  in  more  eulogistic  terms  than  have  so  far  been  used. 


694 


BUSINESS  CYCLES 


whether  our  process  still  persists.  The  method  of  deriving  an  answer  is 
to  locate  the  postwar  period  in  our  cyclical  schema,  to  formulate 
the  expectations  which  follow  from  that,  and  to  see  how  far  they  agree 
with  observed  fact.  According  to  that  schema  the  postwar  time  up  to  the 
world  crisis  covers  parts  of  the  recession  and  depression  phases  of  our 
third  Kondratieff  which  underlie  two  incomplete  Juglars.  If  fluctuations 


1919  1920  1921   1922  1923  1924  1925  1926  1927  1928  1929  1930  1931   1932  1933  1934  1935 
CHART  XL. — German  postwar  "pulse"  (see  Appendix,  p.  1066). 

behaved  as  they  did  before  the  war,  those  Juglars  would  be  the  third  and 
fourth  of  that  Kondratieff.  The  third  would  complete  the  recession,  the 
fourth  would  entirely  lie  on  (but  not  complete)  the  depression  phase  of 
the  latter.  We  ought  to  be  able,  finally,  to  discern  the  Kitchin  wave 
superimposed  on  those  two.  The  time  series  picture  of  all  this  must  then 
link  up  with  the  historical  facts  of  the  industrial  process  behind  it.  Expecta- 
tions are  perfectly  definite  and  will  be  formulated  as  our  picture  unfolds. 
It  is  recommended,  however,  that  the  reader  formulate  them  now  and 


1919-1929 


695 


compare  them  with  the  Postwar  Pulse  Charts  herewith  presented  (Charts 
XXXIX,  XL,  and  XLI). 


•J 

^ 

\ 

^v- 

^/ 

v^/1 

V 

\A 

HOI 

ESA 

.E  PRIC 

s 

vxv 

S^^s, 

/^ 

"V«— 

\ 

r" 

"V 

V 

sT- 

SX^ 

I 

j 

i 

i 

f 

1 
1 

\ 

IN 

'ERE 

STI 

ATE 

i 

1 

1 

\ 

•1 

Iv/ 

T 

V 

j 

«  T" 

y 

r 

"V 

» 

* 

\l 

i 

1 

1 

! 

i    r 

1 

^ 

! 

\ 

i 

1 

Vl 

A/ 

I 

i 

/ 

*wj 

v/w 

^VX 

Dt 
^U 

3OSI 
v>w 

rs<f 

CIR( 

:UUA 

noN 

JvJ 

i 
i 

y^ 

\S*^ 

vX 

V*^ 

LA-/ 

^M 

\x^^ 

^S/ 

^ 

^ 

1 

^ 

If 

V 

f 
^/ 

ROCUCTION  1 

111 

vIDE) 

V 

^ 

V 

V/ 

\f 

i 
i 

vt 

1 

\ 

I 

V 

1 

1 

1919  1920 1921  1922 19231924 1925  1926 1927  1928 1929 1930  1931  1932 19331934  1935 
CHART  XLI. — British  postwar  "pulse"  (see  Appendix,  p.  1067). 

B.  Comments  on  Postwar  Patterns. — Now  we  will  drop,  if  only  for  a 
few  paragraphs,  the  practice,  imposed  upon  us  by  the  nature  of  our  task, 
of  treating  the  institutional  framework  of  society,  the  attitudes  of  indi- 
viduals and  groups,  and  the  policies  resulting  from  a  given  social  pattern 
as  data  of  our  economic  process,  and  changes  in  these  data  as  external 
factors.  We  will  glance  at  the  social  process  as  a  whole  and  in  so  doing 
adopt  the  convenient,  though  possibly  inadequate,  hypothesis  of  Marx- 
ism, according  to  which  social,  cultural,  political  situations  and  the  spirit 
in  which  and  the  measures  by  which  they  are  met,  derive  from  the  working 


696  BUSINESS  CYCLES 

of  the  capitalist  machine.  Our  cyclical  schema  lends  itself  to  this  view, 
not  only  because  of  the  length  of  its  longest  wave,  which  brings  long-run 
social  changes  within  the  reach  of  business-cycle  analysis,  but  also  because 
it  stresses  that  kind  of  economic  change  that  is  particularly  likely  to 
break  up  existing  patterns  and  to  create  new  ones,  thereby  breaking  up  old 
and  creating  new  positions  of  power,  civilizations,  valuations,  beliefs,  and 
policies  which  from  this  standpoint  are,  therefore,  no  longer  "external." 
The  standard  illustration  is  afforded  by  those  innovations  which  drove  the 
artisan's  shop  into  modest  reservations  and,  together  with  the  artisan's 
shop,  also  the  artisan's  world.  Gathering  up  the  threads  that  lie  all  over 
our  edifice — see,  particularly,  Chaps.  Ill,  VI,  and  VII — we  might  thus 
try  to  understand  the  social  configuration  of  the  postwar  period  from 
the  economic  process  we  have  analyzed. 

But  we  should  find  the  task  more  difficult  than  the  analogous  one 
in  the  cases  of  the  first  and  second  Kondratieffs.  There  the  social 
process  and  its  cultural  and  political  complements  were  not  difficult  to 
interpret  in  the  sense  of  the  working  hypothesis  which  for  the  moment 
we  have  adopted.  All  that  was  not  covered  by  it  we  could  comfortably 
stow  away  as  atavisms.  This  is  not  so  in  the  case  of  the  Neomercantilist 
Kondratieff.  If  the  reader  refers  to  Chap.  VII,  Sec.  E,  he  will  find  that 
we  had  to  recognize,  besides  phenomena  that  indicated  consistent  develop- 
ment of  previous  tendencies,  the  presence  of  other  phenomena  that  did 
not  seem  to  fit  into  the  same  current  but  rather  to  fight  against  it:  they 
looked  like  a  revolt  against  the  rational  or  rationalistic  civilization  of  that 
epoch.  Of  course,  it  is  easy  to  label  them,  too,  as  atavisms.  This 
sounds  convincing  in  some  cases,  for  instance,  in  the  case  of  the  German 
legislation  for  the  protection  of  the  artisan  class  (Ilandwerkergesetz, 
1897).  Here  we  see  a  dying  stratum  trying  to  defend  its  crumbling 
basis  by  political  means.  It  is  not  so  convincing  in  others,  and  any 
open  mind  must  admit  the  possibility  that  a  movement  of  such  breadth 
and  depth  may  have  been  more  than  an  atavism  or  the  last  card  of  a 
decaying  class.  The  fact  that  the  writer  had  no  better  name  to  offer  for 
it  than  Neomercantilism  sufficiently  shows  that  so  far  he  has  not  succeeded 
in  interpreting  it  to  his  own  satisfaction.1  For  that  term,  at  best,  gets 
hold  of  one  of  many  aspects  and  is  as  inadequate  as  Nationalism  or  Anti- 

1  The  atavism  theory  was  expounded  by  the  present  writer  in  his  study  entitled  Zur 
Soziologie  der  Imperialismen,  Archiv  fiir  Socialwissenschaft,  1916.  The  Marxist  theory  of 
imperialism  (Bauer,  Hilferding)  mentioned  in  Chap.  VII  made  the  attempt  to  construe 
prewar  imperialism  as  an  outgrowth  of  the  conditions  of  trustified  capitalism.  This 
explanation,  which  conserves  unity  of  principle,  has,  of  course,  great  attractions  for  every 
mind  that  has  an  analytical  bent,  and  could  be  generalized  to  include  postwar  fascism. 
It  is  not  possible  here  to  expound  the  reasons  why  it  is  inadequate.  A  glimpse  of  a  view 
that  now  seems  to  the  writer  to  be  nearer  the  truth  than  either  the  Marxist  or  his  own 
theory  is  embodied  in  Karl  Renner's  concept  of  Social  Imperialism  (Sozialimperialismus). 


1919-1929  697 

rationalism  would  be.  Now  that  tendency  or  attitude  did  not  perish. 
On  the  contrary,  it  developed  during  the  postwar  period  and  in  develop- 
ing revealed  itself  more  fully  in  the  Corporative  or  Totalitarian  or  Fascist 
State  and  also  became  ideologically  articulate.  However  much  the  war 
— and  the  circumstances  of  the  " Have-not  nations" — may  have  had  to 
do  with  concrete  forms,  mechanisms,  timings,  and  surface  events,  that 
departure  from  the  road  that  leads  from  capitalism  to  orthodox  socialism 
is  not  "due"  to  it,  and  the  general  drift  of  this  page  might  have  been  the 
same  had  it  never  occurred.1  The  answer  to  the  question  how  this 
development  may  be  expected  to  affect  our  cyclical  process  depends  on 
the  kind  of  planning  that  a  fascist  government  undertakes :  given  sufficient 
power  and  insight  in  a  central  authority,  innovation  may  of  course  be 
planned  for  in  such  a  way  as  to  minimize  disturbance. 

1.  If  this  component  of  postwar  history  can  be  traced  to  prewar 
sources,  everything  can.  For  the  only  other  component — the  "socialist" 
one — is  perfectly  en  regie  from  the  standpoint  of  our  working  hypothesis 
and  may  readily  be  described  in  terms  of  the  rationalizing,  leveling, 
mechanizing,  and  democratizing  effects  of  capitalist  evolution.  This  is 
too  obvious  to  detain  us,  and  only  a  few  points  of  particular  relevance  to 
our  subject  need  elaboration. 

First,  the  rise  of  the  labor  interest  to  a  position  of  political  power  and 
sometimes  of  responsibility,  which  is  but  the  most  conspicuous  of  the 
symptoms  of  a  profound  change  in  social  structures,  is  clearly  a  product 
of  capitalism  in  our  sense  of  the  term,  which  created  a  political  world  and 
political  attitudes  fundamentally  incompatible  with  itself  even  where,  as 
in  the  United  States,  the  labor  interest  was  (within  our  period)  not 
politically  dominant.  The  habit  of  the  old-fashioned  liberal — in  the 
European  sense  of  the  word — of  blaming  "politics"  for  almost  everything 
he  considers  less  than  satisfactory  in  the  capitalist  world  is,  as  far  as  this 
goes,  in  fact  open  to  the  objection  that  in  blaming  "politics"  he  is 
blaming  a  product  and  an  essential  element  of  the  system  he  approves. 
Taking  the  social  system  of  capitalism  as  a  whole,  it  is  meaningless  to 

1  The  question  naturally  arises  whether  that  turn  of  phrase  is  admissible  at  all.  The 
answer  is  that,  even  from  the  strictest  determinist  standpoint,  there  is  meaning  to  an 
analysis — a  Gedankenexperiment — of  what  difference  the  elimination  of  an  element  in  the 
historical  process  would  make.  Another  problem  may  be  noticed  in  passing:  it  has  been 
held  that  great  wars  are  merely  incidents  of  the  process  of  the  Long  Wave,  "regularly" 
coming  near  its  upper  turning  point  as  a  consequence  of  the  expansion  of  the  productive 
apparatus  which  then  fights  for  markets.  There  is  nothing  in  that  "regularity."  For  it  is 
obviously  absurd  to  interpret  either  the  Napoleonic  or  the  Franco-German  War  as  explo- 
sions of  redundant  capitalist  forces — the  French  industry  was,  in  Napoleonic  times,  to  a 
great  extent  in  the  craft  stage,  and  even  in  1870  Germany  was  a  primarily  agricultural 
country.  The  problem,  therefore,  arises  only  in  the  case  of  the  war  of  1914  to  1918.  The 
writer  thinks,  though  he  cannot  prove  here,  that  the  history  of  its  causation  clearly  yields 
a  negative  result. 


698  BUSINESS  CYCLES 

say  that  it — or  any  element  of  it,  e.g.,  the  gold  standard — is  checkmated 
by  " politics."  What  ought  to  be  said — on  this  level  of  analysis — is  that 
it  checkmates  itself. 

Second,  it  is  worth  mentioning  that  capitalism,  also  by  its  own  work- 
ing, evolves  a  phenomenon  which,  or  the  importance  of  which,  was  not 
foreseen  by  Marx :  the  clerical  class.  The  growth  of  the  laboring  stratum 
proper,  relatively  to  the  increase  of  gainfully  occupied  persons,  ceased 
in  the  first  decade  of  this  century,  but  the  relative  growth  of  the  salaried 
employee  (roughly  equal  to  "white  collar")  then  became  spectacular1 — 
for  obvious  reasons  of  capitalist  technique.  The  interests  of  this  class — 
the  "logic  of  its  situation" — and  its  attitudes  differing  considerably  from 
the  interests  and  attitudes  of  "workmen,"  we  have  here  a  factor  to  the 
power  of  which  much  of  the  politics  and  policies  of  postwar  times  may 
be  traced,  particularly  in  Germany.  We  must  dismiss  this  most  inter- 
esting subject  with  two  remarks:  that  this  New  Middle  Class,  as  it  has 
been  called,  forms  in  some  countries  and  comes  near  to  forming  in 
others,  together  with  farmers  (peasants)  and  small  businessmen  (mainly 
retailers),  a  majority  of  the  population,2  which,  though  split  into  widely 
differing  sections,  yet  feels  and  acts  uniformly  in  many  cases;  and  that 
in  fundamental  attitude  it  is  as  hostile  to  the  interests  of  the  bigger  and 
big  bourgeoisie  as  is  the  working  class  in  the  narrower  sense  of  the  term, 
though  also  hostile  to  the  interests  of  the  latter.  It  is  in  the  light  of 
these  facts  and  not  in  the  light  of  the  simple  but  entirely  unrealistic3 

1  Take,  for  instance,  the  German  case.     The  figures  of  1907  and  1925  are  but  imper- 
fectly   comparable.     But    the    number   of   gainfully   employed    persons    (Erwerbstdtige) 
increased  by  24  per  cent,  the  number  of  workmen  by  24  per  cent  (34  per  cent  in  industry, 
however;  there  was  a  decline  of  9  per  cent  in  agriculture)  and  the  number  of  salaried 
employees  (Beamte  und  Angestellte}  by  66  per  cent,  in  industry  (and  Handwerk)  alone  by 
135  per  cent.     The  absolute  figure  was  in  1925  a  little  over  5.2  millions. 

2  Take,  again,  Germany  in  1925:  "Independents"   (including  the  members  of  their 
families,  whether  working  in  the  business  or  not,  and  the  independents  without  a  profession, 
but  deducting  one  million  for  the  "capitalist  class")  were,  roughly,  23  millions;  employees 
with  their  families,  roughly,  10  millions — #3  millions  in  all.     Laborers  (and  their  families; 
we  include  domestic  servants)  were  not  quite  28  millions  (total  population,  roughly,  62 
millions). 

3  Many  sociologists  and  economists  try  to  keep  to  that  simple  schema  by  means  of  two 
additional  propositions:  first,  that  the  farmer  and  small  businessman  are  on  their  way  to 
extinction,  hence  will  not,  in  the  long  run,  help  to  swell  that  intermediate  stratum;  second, 
that  the  white-collar  employee  is  really  a  proletarian  like  the  manual  laborer  and  ought  to 
take  his  true  place  within  the  proletarian  front.     We  need  not  go  into  the  question  whether 
or  not  these  propositions  will  turn  out  true  in  the  long  run.     For  our  purpose  it  is  sufficient 
to  state  that  they  are  not  true  for  as  much  of  the  postwar  period  as  has  so  far  elapsed, 
Ex  visu  of  that  span  of  time,  it  is,  in  particular,  not  true  to  say  that  the  bulk  of  white- 
collar  employees  has  refrained  from  joining  the  proletarian  front  merely  because  of  the 
pressure  to  which  they  are  exposed.     This  assertion  about  pressure  seems  to  rest  on  a  false 
generalization  from  inadequate  observation.     It  is  not  so  easy  to  exert  pressure  of  the  kind 


1919-1929  699 

contrast  between  property  owners  and  proletarians  that  postwar  patterns 
must  be  understood. 

Third,  capitalist  evolution  not  only  upsets  social  structures  which  pro- 
tected the  capitalist  interests,  by  progressively  eliminating  precapitalist 
strata  from  politics  and  public  administration  and  by  creating  new  posi- 
tions of  political  power,  but  also  undermines  the  attitudes,  motivations, 
and  beliefs  of  the  capitalist  stratum  itself.  Even  if  an  industrial  family 
happens  to  own  a  given  concern,  wholly  or  nearly  so,  and  if  its  members 
actually  manage  it,  they  do  not  under  modern  conditions  look  upon  it  in 
the  way  industrial  families  used  to  do  in  the  past.  Their  attitude  is 
more  distant,  less  personal,  more  rationalized.  But  the  leading  men  of 
the  giant  concerns  as  a  rule  fill  a  specialized  function  in  a  spirit  which 
resembles  that  of  the  employee  properly  so  called,  and  tend  to  distinguish 
between  their  success  and  that  of  the  concern,  let  alone  that  of  the  share- 
holders. Moreover,  the  loosening  of  the  family  tic — a  typical  feature  of 
the  culture  of  capitalism — removes  or  weakens  what,  no  doubt,  was  the 
center  of  the  motivation  of  the  businessman  of  old.  Finally,  the  top 
group — say,  40,000  men  and  their  families  in  this  country  and  just 
about  as  many  in  Germany — absorbs,  subconsciously  and  by  an  infinite 
number  of  channels,  views,  habits,  valuations — cultural  worlds — that 
are  not  its  own.  "Capitalists"  cease  to  believe  in  the  standards  and 
moral  schemata  of  their  own  class.  They  adopt,  or  connive  at,  many 
things  which  their  predecessors  would  have  considered  not  only  injurious 
to  their  interests  but  dishonorable:  in  surveying  modern  economic  fact, 
one  cannot  but  be  struck  by  the  discovery  of  how  much  of  the  typical 
behavior  of  the  bourgeoisie  of  the  nineteenth  century  was  extra-economi- 
cally  conditioned.  All  this,  of  course,  links  up,  in  a  way  that  hardly 
requires  explanation,  with  the  decrease  in  the  importance  of  the  entre- 
preneurial function  noticed  in  Chap.  Ill,  Sec.  C,  5  and  Chap.  IV,  Sec.  B. 

Fourth,  both  the  rise  to  power  of  strata  untinged  by  bourgeois  atti- 
tudes and  the  fact  that  these  attitudes  lose  their  hold  on  the  bourgeois 
stratum  itself,  which  to  an  increasing  extent  allows  itself  to  be  educated 
by  its  new  masters  and  the  intellectual  exponents  of  these  masters — as  it 
used  to  allow  itself  to  be  educated  by  its  former  feudal  masters — combine 
to  produce  the  anti-saving  attitude  of  our  epoch  so  clearly  voiced  in  its 
popular,  as  well  as  scientific,  literature  on  economic  theory  and  policy. 
Saving  with  a  view  to  providing  revenue  for  an  indefinite  family  future 
was  part  not  only  of  the  economic  but  also  of  the  moral  scheme  of  life  of 
the  typical  bourgeois.  The  attempt  to  prove  that  such  thrift  is  harmful 
to  the  interest  of  the  masses  always  has  been  a  major  element  in  anti- 
capitalist  arguments,  which  without  it  would  be  open  to  a  dangerously 

contemplated  on  a  body  of  employees.  And  nothing  can  be  clearer  than  the  fact  that  it  is 
precisely  the  proletarian  class  consciousness  that  is  wanting  in  that  stratum. 


700  BUSINESS  CYCLES 

obvious  reply.  Attempts  to  prove  that  it  is  also  harmful  to  the  capitalist 
interest  itself  have  never  been  wanting.  But  in  our  time  the  lesson  is 
being  learned  and  beginning  to  motivate  public  policy.  Whatever  its 
merits  or  demerits,  its  success  must  be  understood  as  part  of  a  general 
short-run  attitude  of  modern  man  toward  economic  problems  and  situa- 
tions, which  follows  from  the  changes  in  social  structure.  All  this  spells  a 
profound  change  in  the  environment  within  which  the  capitalist  engine 
works,  and  in  particular  may  invalidate  in  a  near  future  that  extrapola- 
tion of  past  performance  which,  despite  the  arguments  put  forth  by  some 
authors  in  order  to  establish  a  tendency  toward  retardation,  it  was  thought 
justifiable  to  make  (Chap.  IX,  Sec.  B;  also  see  infra,  Chap.  XV,  Sec.  H). 
2.  But  recognizing  thus  fully  the  relation,  at  least  of  interdependence, 
possibly  of  causation,  which  exists  between  capitalist  evolution  and  its 
social  and  sociopsychological  complement,  no  more  debars  us  from 
recognizing  the  existence  of  distinguishable  spheres  of  social  activity 
between  which  in  a  given  case  given  effects  may  be  apportioned,  than 
recognition  of  universal  interdependence  of  prices  debars  us  from  dis- 
tinguishing them  and  from  tracing  given  effects  to  the  behavior  of,  say, 
one  of  them.  Moreover,  every  such  sphere,  however  much  the  product 
of  one  comprehensive  process,  acquires,  when  once  formed,  a  Jife  and 
mechanism  of  its  own  that  enjoys  many  degrees  of  freedom.  This  is 
enough  to  justify  our  going  on  to  work  with  the  concept  of  External 
Factors.  In  particular,  it  is  clear  that  we  cannot  from  a  study  of  eco- 
nomic conditions  alone  determine  what  will  happen  in  the  political 
sphere,  and  if  we  cannot,  there  is  not  much  practical  value  in  emphasizing 
any  quasi-religious  belief  in  such  determinateness  that  we  may  harbor. 
On  the  contrary,  we  must  deal  with  all  the  facts  of  each  sphere  as  we  find 
them — which  precisely  means  that  they  are  external  factors  for  one 
another.  For  instance,  it  is  not  enough  and  it  is  perhaps  not  even  very 
enlightening  to  know  that — possibly — England's  economic  conditions  and 
interests  to  some  extent  explain  her  general  attitude  to  the  United  States 
during  the  Civil  War.  This  will  not  explain  why  she  came  within  an  ace 
of,  yet  refrained  from,  interfering  by  the  force  of  arms.  Even  the 
necessity — so  repulsive  to  many  students — of  taking  account  of  the 
personal  element  cannot  be  denied.  Hence,  purely  economic  diagnosis 
of  the  type  for  which  political  action  is  merely  a  "disturbing  factor"  is 
not  necessarily  devoid  of  meaning  but  may  be  instructive  to  bourgeois  and 
socialist  alike  (see  ante,  Chap.  I).  We  may  also  note  in  passing  that 
there  is  no  justification  whatever  for  the  habit  some  economists  have  of 
skirting  uncomfortable  issues  by  pointing  to  the  compelling  force  of  social 
and  political  situations.  To  any  situation  the  political  mechanisms  and 
their  crews  can  react  in  many  different  ways.  What  is  compulsion  to 
some  is  not  compulsion  to  others.  In  any  case,  it  is  not  his  professional 


1919-1929  701 

equipment  that  entitles  the  economist  to  pronounce  about  this  question. 
And  the  truth  or  falsity  of  his  economic  propositions  is  as  independent  of 
"political  necessity"  as  the  truth  or  falsity  of  a  doctor's  diagnosis  is 
independent  of  whether  or  not  the  patient  is  willing  or  able  to  act  upon  it. 

In  this  sense  the  World  War  is  for  us  an  external  factor.  From  the 
facts  the  above  remarks  were  intended  to  cover,  it  seems  to  follow  that 
it  did  not  "create"  any  of  the  fundamental  social  features  of  the  postwar 
world,  although  it  accentuated  some  and  may  have  anticipated  others. 
The  physical  destruction — including  the  expenditure  of  productive  energy 
on  that  tremendous  "excess  in  consumption" — and  the  loss  of  life  in  the 
most  active  age  groups  were  made  up  quickly,  the  former  with  a  prompt- 
ness which  in  another  social  atmosphere  would  have  been  admired  as  a 
marvel  of  industrial  efficiency.  All  this  reduces  for  our  purpose  to  two 
statements:  first,  that  physical  destruction,  reinforced  by  the  accumula- 
tion of  omitted  replacements  and  investments,  became  the  source  of  a 
reconstruction  demand,  not  only  in  our  three  or  in  all  belligerent  coun- 
tries but  also  in  others,  which  accentuated  the  prosperities  and  revivals 
that  occurred  up  to,  roughly,  the  middle  twenties,  although  this  effect 
fully  asserted  itself  only  in  this  country  and  was  in  England  and  Germany 
counteracted  or  delayed  by  the  well-known  circumstances  to  be  noticed 
presently;  second,  that  the  shift  from  war  to  peace  conditions  involved 
rearrangements  that  almost  wholly  account  for  the  short  "jolt"  in  1918 
and  partly  for  the  crisis  of  1921,  which  was,  however,  in  step  with  the 
ordinary  run  of  cycles  and  was  only  intensified  by  this  factor  (as  was  the 
crisis  of  1815). 

The  moral  disorganization  brought  about  by  the  war  and  by  inflation 
and  their  effects  on  the  cultural  inheritance  of  mankind  would,  in  any  social 
history  of  our  time,  have  to  occupy  a  much  more  prominent  place.  It 
accounts  to  this  day  for  many  of  the  most  striking  phenomena  in  all 
sectors  of  social  life,  and  it  was  mainly  responsible  for  the  lack  of  stamina 
that  was  displayed  by  the  ruling  strata  in  some  countries  and  that  sud- 
denly forced  issues  into  practical  politics  for  which  evolution  was  pro- 
viding, but  had  not  yet  provided,  the  necessary  conditions.  One  example 
will  suffice  to  illustrate  this  proposition  and  its  bearing  on  our  subject: 
no  serious  socialist  would  in  1914  have  expected — no  serious  socialist 
would  have  desired — that  practical  realization  of  the  socialist  program 
would  by  1918  become,  in  many  European  countries  and  in  an  otherwise 
quite  immature  situation,  a  political  issue,  merely  because  of  the  moral 
breakdown  of  the  ruling  strata.  When  socialist  parties  were,  neverthe- 
less, confronted  by  that  fact,  they  were  unwilling  or  unable  to  use  the 
Russian  method  and  to  bear  down  resistance  in  streams  of  blood  in  order 
to  force  into  the  socialist  mould  a  humanity  that  would  not  shape  for  it 
of  its  own  accord.  Deadlock  ensued  in  which  neither  "capitalist"  nor 


702  BUSINESS  CYCLES 

"socialist"  policy  prevailed.  As  a  sort  of  compromise,  the  capitalist 
engine  was  allowed  to  work,  but  was  put  under  a  pressure  that  prevented 
it  from  working  according  to  design.  We  shall  presently  glance  at  the 
case  of  Germany,  which  exemplifies  that  state  of  things  particularly  well. 
But  its  fundamental  contours  also  show  elsewhere,  for  instance  in 
England.  It  is  not  generally  realized — many  people  have  a  strong  aver- 
sion to  realizing1 — how  important  this  is  for  an  explanation  of  the  eco- 
nomic history  since  the  war  and  how  many  difficulties  were  created  or 
increased  by  the  attempt  to  handle  the  essentially  intracapitalistic 
problems  of  this  period  by  anticapitalistic  methods,  which  yet  nowhere, 
with  the  exception  of  Russia,  went  far  enough  to  replace  the  economic 
forces  they  weakened  by  those  of  another  social  system.  Short  of 
socialization,  the  most  obvious  measure  of  the  pressure  to  which  the 
capitalist  organism  has  been  and  is  being  exposed,  is  the  amount  and  the 
progressiveness  of  taxation.2  Independently  of  this,  the  direct  effects  on 
the  economic  process  of  income,  corporation,  and  inheritance  taxes  are 
considerable  (see  below,  Sec.  C).  For  both  reasons,  a  prominent  place 
will  be  allocated  to  fiscal  policy  in  the  analysis  that  is  to  follow.  In  con- 
sequence, other  aspects  of  domestic  policies  will  have  to  be  neglected. 

3.  Thus,  the  war  undoubtedly  precipitated,  and  gave  their  particular 
forms  to,  developments  which  it  is  reasonable  to  assume  would  have  come 
about  without  it,  but  would  have  come  about  more  slowly  and  in  different 
forms.  It  is  in  connection  with  these  developments  that  foreign  policies 
and  the  problems  of  international  economic  relations  in  general  must  be 
seen.  Their  history  naturally  divides  up  into  three  periods :  the  period 
of  continuing  economic — in  some  points  even  extraeconomic — warfare, 
from  the  armistice  to  1924  (London  conference;  Dawes  plan);  the  period 
during  which  the  arrangements  arrived  at  were  widely  believed  in  and 
acted  upon,  and  which  lasted  from  the  Dawes  plan  to  roughly  the  end  of 
1927;  and  the  period  of  increasing  friction,  ending  in  the  midst  of  the 
world  crisis  in  the  Hoover  moratorium  and,  later  on,  in  the  liquidation, 
viafacti,  of  the  reparations  and  inter- Allied  debts.  Since  the  first  period 
coincides  with  German  postwar  inflation,  which  we  exclude,  all  that  need 
be  said  about  it  is  that,  while  American  business  situations  were  for  the 

1  That  aversion  is  easy  to  understand.     Most  people  who  write  and  think  about  postwar 
events  are  in  sympathy  with  socialism  of  some  sort.     But  most  of  them  are  not  militant, 
so  far  as  the  ultimate  goal  is  concerned.     It  is  natural  for  such  essentially  "transitional" 
mentality  to  accept  for  the  time  being  a  highly  socialized  capitalism  as  a  substitute.     It 
is  no  less  natural  that,  having  accepted  it,  they  try  to  believe  in  its  possibilities  and  to 
fight  against  any  suggestion  that  it  may  be  productive  of  consequences  which  from  their 
own  standpoint  they  would  have  to  consider  as  "undesirable." 

2  It  also  affords,  excepting  times  of  national  emergency,  the  most  obvious  measures  of 
the  political  power  of  the  capitalist  class.    Everything  else  may  be  phraseology;  the 
income  and  inheritance  taxes  are  facts. 


1919-1929  703 

time  being  little  and,  if  anything,  favorably  affected  by  the  course  of 
international  events,  England  suffered  greatly:  her  exports  to  Germany, 
for  instance,  were  even  after  1924  smaller  by  about  one-third  than  they 
had  been  in  1913  (measured  in  the  1925  units  of  the  Statistische  Reichs- 
amt),  but  before  that  year  her  troubled  relations  to  her  "best  customer*' 
certainly  helped  to  account  for  her  share  in  the  depression  of  1921  and  for 
he  unsatisfactory  recovery  that  followed. 

As  regards  the  second  period,  our  statement  that  the  arrangements 
arrived  at  were  acted  upon  requires  qualification.  The  world  never 
squarely  faced  their  fundamental  consequences,  the  unavoidable  Ger- 
man exports  in  particular,  but  poured  American  and  other  credits  on  all 
the  unsolved  problems  and  plastered  gold  currencies — most  curious 
atavisms  though  they  were,  in  a  world  otherwise  resolved  not  to  play  the 
capitalist  game — on  essentially  untenable  situations.  Since  it  is  impossi- 
ble to  enter  upon  these  matters  and  since  the  salient  facts  may  be  assumed 
to  be  familiar,  we  can  for  our  purpose  compress  the  comments  that  would 
be  necessary  into  the  statement  that  this  "export  of  American  capital" 
did  not  act  as  we  would  expect  capital  export  to  act  under  more  normal 
conditions,  but  that  for  the  time  being  it  largely  counteracted  the  dis- 
turbance that  would  otherwise  have  ensued  from  international  political 
payments.  Moreover,  while  it  thus  relieved  stringencies  in  many  coun- 
tries, it  did  not  create  corresponding  ones  in  the  lending  country,  because 
America  still  remained  a  creditor  on  current  account,  even  apart  from 
the  short  balances  which  for  familiar  reasons  continued  to  flow  toward  her. 
This  explains  the  astonishing  fact  that  those  "political  transfers"  will 
not  play  any  great  role  in  our  analysis  of  the  processes  of  that  period 
except  in  the  German  case,  and  that  even  in  the  latter  case  it  is  transfers 
to  Germany,  rather  than  transfers  from  Germany,  that  shaped  situations. 
This  even  applies  to  the  third  period,  almost  until  the  outbreak  of  the 
world  crisis  when,  though  partly  for  political  reasons,  short  balances  took 
fright.  But  the  fundamental  facts  of  the  situation,  accentuated  as  well 
as  expressed  by  protectionist  tendencies  and  the  passing  of  the  Locarno 
atmosphere,  had  begun  to  assert  themselves  before  that. 

It  should  be  observed,  however,  that  if  those  temporary  solutions  of 
the  international  financial  problems  created  by  the  war  proved  inade- 
quate, and  if  their  economic  consequences,  including  their — secondary — 
share  in  the  causation  of  the  world  crisis,  turned  out  as  they  did,  this  was 
only  because  of  the  political  environment  within  which  they  had  to  work, 
i.e.,  that  our  assertion  of  their  inadequacy  must  be  understood  to  be 
relative  to  the  social  situation  previously  glanced  at.  They  were  bankers' 
solutions,  which  the  nations  concerned  were  unwilling  to  accept  and 
which  they  defeated  by  refusing  to  allow  the  mechanisms  to  work  on 
the  functioning  of  which  the  authors  of  those  solutions  relied.  Looked 


704  BUSINESS  CYCLES 

at  as  business  propositions,  they  would  not,  in  a  peaceful  world  accepting 
bourgeois  standards,  have  been  obviously  absurd,  and  it  would  not  in 
such  a  world  have  been  unreasonable  to  expect  that  they  would  work  out 
in  the  end.  The  well-meant  proposals  of  all  those  international  con- 
ferences that  met  to  discuss  the  gradual  removal  of  trade  barriers  look  to 
us  like  strange  anachronisms  and  certainly  were  as  futile  as  proposals  for 
disarmament  must  be  in  a  world  in  which  every  nation  that  counts  is  bent 
on  armament.  But  they  were  perfectly  sound  economics.  JEven  the 
gold  currencies  were  such  failures  only  because  trade  barriers,  fiscal 
policies,  social  and  military  expenditure,  and  insistence  on  higher  money 
wages  did  not  allow  them  to  function  and  because  in  that  hostile  environ- 
ment short  capital  was  rushing  about  like  a  hunted  hare.  Given  all 
these  facts,  it  was  and  is  indeed  little  short  of  ridiculous  to  trust  to  the 
remedial  forces  of  laixser  faire.  But  since  they  do  not,  any  more  than 
the  war  itself  and  Versailles,  uniquely  follow  from  the  logic  of  our  evolu- 
tionary process,  it  is  not  to  the  interest  of  clear  thinking  to  speak  of  any 
inherent  tendency  of  capitalism  to  run  into  such  deadlocks.1 

4.  We  will  mention  briefly  a  few  examples  of  various  other  types  of 
war  effects.  In  some  countries,  such  as  New  Zealand,  war  demand  had 
induced  expansions  which  were  clearly  untenable  under  normal  conditions. 
Even  disregarding  the  speculative  excesses  that  actually  occurred,  we 
have  no  difficulty  in  understanding  that  liquidation  was  unavoidable. 
In  New  Zealand  it  began  in  1921  and  continued,  witness  the  abnormal 
number  of  bankruptices,  almost  without  interruption  right  into  the 
world  crisis.  Protection,  borrowing,  and  prosperities  in  many  parts  of  the 
world  mitigated  the  process,  which  up  to  the  crisis  spelled  additional  busi- 
ness for  England — the  total  imports  of  New  Zealand  more  than  doubled 
from  1919  to  1920  and  then  remained  on  a  lower,  but  still  very  high  level 
— and  in  the  crisis  additional  embarrassment  through  frozen  credits. 

1  The  reader  need  not  fear  that  the  above  argument  proceeds  from  any  intention  to 
"whitewash"  capitalism.  For,  as  pointed  out  before,  it  leaves  him  free  to  condemn  as 
much  as  he  pleases  the  social  system  as  a  whole  which  produces  such  deadlocks.  But  it 
serves  no  purpose  to  blur  the  diagnosis  of  that  situation  by  faulty  economic  argument. 
This  has  been  done  not  only  by  those  writers  who  are  prisoners  of  a  rigid  formula,  but  also 
by  economists  who  care  more  about  conveying  their  practical  advice  than  about  analysis. 
The  discussion  on  the  reparation  problem  may  serve  as  an  example.  Instead  of  stressing 
the  simple  and  trivial  truth  that  the  Dawes  tribute  was  morally  inacceptable  to  Germany, 
and  economically  inacceptable  to  the  recipient  countries — which  would  never  have  sub- 
mitted to  having  their  industries  crippled  by  the  German  exports  that  would  have  been  the 
necessary  consequence  of  actual  payment — many,  especially  English,  writers  tried  to  prove 
the  economic  impossibility  of  those  exports  on  grounds  of  elasticity  of  demand,  while  it  is 
perfectly  clear  that,  looked  at  as  a  business  proposition,  the  Dawes  tribute  would  have  been 
nothing  else  but  a  "commission"  paid  by  Germany  for  the  industrial  conquest  of  the 
better  half  of  the  world.  If  the  world  was  unable  to  act  upon  this  possibility,  it  may  have 
been  sound  advice  to  cancel  both  reparations  and  inter-Allied  debts,  but  this  does  not 
save  the  reasons  by  which  that  advice  was  supported. 


1919-1929  705 

The  example  is  really  typical  of  many  more  important  cases,  all  of  which 
contributed  their  special  crises  to  the  world's  economic  situation  after 
1929.  It  is  worth  mentioning  that  in  the  maladjustments  revealed  by  the 
cessation  of  war  conditions,  the  New  Zealand  wage  level  cannot  have 
counted  for  much.  If  the  official  index  may  be  trusted,  real  wages  moved 
under  the  1914  level  until  1928.  This  also  applies  to  other  cases.  The 
widespread  opinion,  for  instance,  that  Australian  wages  were  a  major 
cause  of  subsequent  troubles  is  hardly  borne  out  by  observable  facts. 

In  other  countries  the  war  quickened  industrial  expansion,  which 
it  is  reasonable  to  assume  would  have  ensued  anyhow  in  the  course  of 
time,  and  thus  contributed  to  that  decline  of  the  position  of  Northwestern 
Europe  in  the  world's  economy,  which  in  itself  is  not  a  war  effect.  The 
rise  of  a  native  capitalism  in  the  tropics  is  a  case  in  point.  The  war 
profits  of  India,  for  instance,  gave  an  impulse  to  native  industry  which, 
since  it  directly  affected  Lancashire,  may  be  listed  among  the  factors 
which  shaped  England's  economic  history  during  the  period.  But  the 
effects  of  all  the  other  cases  of  this  type  that  might  be  mentioned — Japan's 
great  stride  in  industrialization,  for  instance— either  dispersed  themselves 
over  too  wide  a  surface  to  signify  for  our  purpose  or  were,  for  the  time 
being,  too  much  compensated  for  by  the  demand  for  foreign  goods — 
investment  goods;  a  characteristic  change  occurred,  as  compared  with 
prewar  times,  in  the  composition  of  European  exports — incident  to  the 
early  stages  of  the  process. 

Still  another  type  of  consequences  may  be  instanced  by  the  breakdown 
of  Russia  and  the  subsequent  developments  under  the  Bolshevist  rule. 
It  is  obvious  that  these  events,  had  they  occurred  in  an  otherwise  "nor- 
mal" world,  would  have  greatly  affected  conditions  in  general  and  the 
cyclical  process  in  particular  in  several  other  countries,  particularly  in 
Germany  and  England.  As  it  was,  however,  the  effect  was  lost  in  a 
mass  of  much  greater  dislocations  of  industry  and  trade,  and  no  distinct 
process  of  adaptation  to  the  new  state  of  things  in  Russia  was  required. 
Business  all  over  the  world  had,  of  course,  partly  to  go  without  the 
stimulus  that  Russian  reconstruction  and  expansion  would  have  offered 
under  other  circumstances:  if  we  extrapolate  her  industrial  development 
in  the  16  years  preceding  the  war,  we  have  no  difficulty  in  forming  an 
idea  about  the  quantitative  importance  of  this  lucrum  cessans.  More- 
over, without  the  loss  of  French  investments  in  Russia,  understandings 
about  reparations  and  inter-Allied  debts  would  have  been  much  easier. 
But  it  should  not  surprise  us  that  we  find  little  of  positive  effects.  Much 
the  same  argument  applies  to  the  case  of  China  and  to  others. 

5.  Since  the  role  which  postwar  protectionism  played  in  the  develop- 
ments of  the  twenties  and  in  the  causation  of  the  world  crisis  has  been  so 
completely  overlooked  by  some  students  of  the  business  cycle  and  so 


706  BUSINESS  CYCLES 

obviously  exaggerated  by  others,  it  will  be  convenient  to  restate  explicitly 
the  view  adopted  for  the  purposes  of  our  analysis — which,  it  should  be 
borne  in  mind,  excludes  the  wider  aspects  of  the  matter,  such  as  the  rela- 
tion of  protectionist  policy  to  human  welfare  and  to  peace.  In  the  first 
years  after  the  war,  duties,  prohibitions,  quotas,  and  other  weapons  from 
the  arsenal  of  protectionism  were,  of  course,  elements  of  the  general 
scheme  of  continuing  economic  war.  But  they  were  also  something  else. 
Adaptation  of  industry  and  trade  to  the  new  conditions  above  referred  to, 
permanent  and  transient,  was  at  best  a  difficult  task,  involving*  in  many 
cases  abrupt  dislocations.  This  becomes  obvious  if  we  glance  at  the 
postwar  figures  both  of  commodities  produced  and  of  commodities 
internationally  exchanged.  In  some  of  these  cases,  protective  tariffs  or 
even  prohibitions  were,  if  not  the  only  means,  yet  the  most  obvious 
means  of  averting  sectional  catastrophes,  from  which  cumulative  proc- 
esses (spirals)  might  easily  have  ensued.  Unequal  depreciation  of  cur- 
rencies, of  course,  greatly  added  to  this  class  of  difficulties.  Many 
measures,  such  as  the  McKenna  duties  and  even  the  Fordney-McCumber 
Act,  must,  in  part  at  least,  be  interpreted  in  this  light  and  on  balance 
probably  mitigated  many  more  difficulties  than  they  created.  At  the 
other  end  of  our  period,  immediately  before  and  during  the  world  crisis, 
a  similar  argument  applies,  especially  after  the  newly  established  cur- 
rencies had  begun  to  give  way,  although  with  lesser  force  and  although 
a  panicky  policy  of  protection  and  "  incapsulation "  then  went  to  obvi- 
ously irrational  lengths. 

In  the  years  from,  roughly,  1924  to,  roughly,  1928  some  steps  toward 
freer  trade  were  actually  made,  some  countries  removing  certain  barriers 
and  the  tariffs  of  others  being  automatically  lowered,  in  cases  of  specific 
duties,  by  depreciation  not  always  compensated  for  by  a  gold  clause. 
However,  it  is  understandable  that  not  more  was  accomplished:  disloca- 
tions and  untenable  war  growths  continued  to  exist;  unequal  depreciation 
of  currencies  was  replaced  by  unequal  stabilization,  which  in  some  cases 
overvalued  and  in  others  undervalued  the  legal  tender  unit;  political 
payments,  especially  but  not  only  in  the  case  of  Germany,  provided  a 
motive  for  aiming  at  an  active  balance  of  trade  entirely  justifiable 
even  from  a  free-trade  standpoint.  There  was,  indeed,  "nationalism." 
Its  outstanding  manifestations  are  to  be  found  in  the  policies  of  the 
majority  of  the  newly  created  small  states  that  tried  to  foster  industrial 
development  at  any  price.  But  that  great  movement  of  which  we 
primarily  think  when  speaking  of  modern  nationalism,  and  which  has 
been  recognized  above  as  allied  to  one  of  the  two  great  components  of 
the  social  atmosphere  of  today,  has  really  little  to  do  with  the  commercial 
policy  of  the  great  nations — and  the  small  ones  of  old  standing — during 
those  years.  That  policy  was  dominated  by  current  vicissitudes  partic- 


1919-1929  707 

ularly  in  England,  America  and — until  the  National  Socialist  party  rose 
to  power — in  Germany.  England  substantially  adhered  to  the  principle 
of  Free  Trade  until  1932.  The  edge  of  German  protectionism  was  in  her 
agrarian  situation.  And  again  it  is  reasonable  to  think  that  on  balance 
and  in  spite  of  many  ill-conceived  individual  measures,  European  pro- 
tectionism during  those  years  hardly  accentuated  prosperities  and 
depressions,  or  created  on  balance  additional  maladjustments  of  major 
importance,  whatever  we  might  have  to  say  about  it  from  other  stand- 
points. Moreover,  it  must  not  be  forgotten  that  the  volume  of  inter- 
national trade  could  in  any  case  not  have  been  expected  to  develop 
along  prewar  trends,  and  that  a  certain  amount  of  "autarky"  was  the 
unavoidable  consequence  of  technological  progress  in  many  branches  and 
countries  (see  Chap.  XIII,  Sec.  C).  Electricity  and  chemistry  alone 
would  have  been  sufficient  to  produce  a  tendency  in  that  direction,  so 
that  the  actual  course  of  events  cannot  be  simply  attributed  to  protec- 
tionist barriers. 

This  country's  famous  "refusal  to  accept  its  creditor  position'*  still 
remains.  In  that  phrase  there  is,  no  doubt,  some  element  of  truth. 
But  since,  as  pointed  out  above,  American  capital  export  to  Europe — it 
reached  about  5  billion  dollars  by  the  end  of  our  period— more  than 
sufficed  to  service  her  claims,  the  consequences  for  the  time  being  were 
merely  to  quicken  reconstruction  in  Europe.  It  was  not  this  mechanism 
that  produced  the  world  crisis,  but  the  world  crisis  that  caused  its  break- 
down. Then,  of  course,  the  situation  thus  created  became  a  major 
factor  in  the  ensuing  depression.  But  even  then  it  is  not  easy  to  see  how, 
had  a  reduction  of  import  duties  been  passed  instead  of  the  Hawley- 
Smoot  Act,  this  could  have  improved  short -run  conditions  in  Europe 
without  aggravating  them  in  America.  Whatever  the  merits  of  free- 
trade  sermons,  they  can  only  apply  either  to  the  course  of  action  that 
might  have  been  followed  had  the  crisis  not  occurred,  or  else  to  the  course 
of  action  that  might  have  been  followed  after  it  had  passed.  With  this 
we  are  not  concerned.  For  our  purpose  it  suffices  to  conclude  that  pro- 
tectionism as  such  played  but  a  minor  role  in  the  cyclical  process  of  the 
postwar  epoch,  and  to  cast  a  glance  at  this  country's  international  balance 
of  1928,  since  that  is  the  last  complete  year  of  the  "prosperity  plateau." 

Commodity  exports  to  Europe  were  then  $2,342,000,000;  imports  of 
commodities  and  services  from  Europe  plus  remittances  of  immigrants 
and  tourists'  expenditure,  very  roughly — estimates  of  tourists'  expendi- 
tures are  particularly  unreliable — 2  billions.  The  resulting  net  credit 
of  between  300  and  400  millions  has  to  be  increased  by  payments  received 
on  war  debt  account — 200  millions — and  net  receipts  from  interest  and 
dividends(P).  This  makes  about  600  millions,  which  must  have  been 
mainly  "paid"  from  additional  credits,  since  the  total  of  monetary  gold 


708  BUSINESS  CYCLES 

in  the  United  States  had  fallen  both  in  1927  and  1928.  It  is  true  that 
the  immigration  of  short  funds,  then  approaching  their  1929  peak,  com- 
plicated the  situation.  But,  with  due  respect  for  the  excellent  motives 
behind  many  of  the  exaggerations  of  which  economists  of  all  countries 
were  guilty  in  the  matter  of  American  policy — Europe's  willingness  to 
lecture  was  more  obvious  than  her  willingness  to  pay — it  is  presumably 
safe  to  say  that  a  sum  of  that  order  of  magnitude  could  not,  without  a 
crisis,  itself  caused  by  other  factors,  have  created  an  unmanageable  prob- 
lem. In  the  ordinary  course  of  things,  adjustments  of  the  commodity 
balance  amounting  to  a  sum  of  the  order  of  300  millions  would  have  been 
possible,  even  in  a  protectionist  world:  exports  alone  could  have  been 
gradually  reduced  to  that  extent  without  serious  repercussions  on  the 
American  situation,1  while  reinvestment  could  have  absorbed  the  rest. 
Again,  it  was  the  crisis  that  prevented  such  adjustments,  and  suddenly 
made  an  insoluble  problem  of  what  otherwise  was  not  only  not  beyond, 
but  on  the  way  to,  a  solution  which,  though  in  the  future  it  might  have 
entailed,  did  not  then  presuppose  free  trade. 

C.  Further  Comments  on  Postwar  Conditions  in  Our  Three  Coun- 
tries.— 

1.  More  nearly  than  any  other  country,  the  United  States  displayed, 
and  substantially  retained  until  the  world  crisis,  a  frame  of  mind  appro- 
priate to  the  task  of  running  the  capitalist  machine,  even  to  the  extent 
of  reducing  what  was  almost  universally  disapproved  of  as  an  "un-Ameri- 
can radicalism*'  to  still  smaller  importance  than  it  had  had  before  the 

1  This  reduction  would  have  been  brought  about  automatically  by  reduced  capital 
exports  if  the  meaningless  policy  of  fostering  commodity  exports  had  been  discontinued. 
Some  planning  would,  however,  have  been  necessary  in  order  to  safeguard  those  sectors  on 
which  the  reduction  of  commodity  exports  would  primarily  have  impinged.  For  it  must 
be  remembered  that  while  American  exports  did  not  greatly  signify  as  compared  with  the 
total  of  the  nation's  business,  they  did  signify  for  agriculture,  textiles,  metals,  machinery, 
and  vehicles.  It  should  also  be  remembered  that  we  are  discussing  relations  with  Europe 
only.  The  relations  of  this  country  with  other  parts  of  the  world  raise  different  problems, 
which  cannot  be  discussed  here.  It  is  obvious,  however,  that  as  far  as  the  period  under 
discussion  is  concerned  the  analogy  with  the  English  case  in  the  second  half  of  the  nine- 
teenth century  is  misleading,  because  England's  established  creditor  position  then  bore  a 
different  relation  to  the  additions  that  were  made  to  it  in  any  given  year.  A  protectionist 
policy  would,  therefore,  have  created  many  more  difficulties  in  her  case  than  it  can  have 
created  in  the  case  of  the  United  States :  while  a  creditor  position  is  rapidly  being  built  up, 
the  lending  country  tends  to  be  in  the  position  of  debtor.  Needless  to  say,  nothing  of 
this  is,  or  intended  to  be,  an  argument  for  protection  in  general.  But  one  more  error 
about  this  country's  creditor  position  should  be  mentioned,  viz.t  the  error  that  it  was  due  to 
the  war  and  would  not  have  developed  without  it.  This  is  true  only  of  the  11  billions  of 
"political"  claims  and  beyond  them  for  the  speed  of  the  process.  But  fundamentally 
the  United  States  were  well  on  the  way  out  of  a  debtor  nation's  and  toward  a  creditor 
nation's  position  in  the  last  prewar  decade,  and  it  is  reasonable  to  assume  that  they  would 
in  any  case  have  reached  the  latter  by  1920. 


1919-1929  709 

war.  Such  deviations  as  occurred  from  those  principles  of  action  that 
are  associated  with  the  logic  of  the  capitalist  process  were  due,  rather  than 
to  the  intrusion  of  ideas  hostile  to  that  logic,  to  the  failure  to  adapt  old 
ideas  to  the  new  situation,  as  exemplified  by  so  much  as  there  was  in  the 
"refusal  to  accept  the  implications  of  this  country's  new  creditor  posi- 
tion" just  discussed. 

But  apart  from  this  and  possibly  monetary  policies  (see  infra,  Sec.  F) 
and  the  rate  of  municipal  expenditure,  there  were  no  lesions  inflicted  on 
the  system  by  action  from  the  political  sphere.  On  the  contrary,  while 
the  nation  was  bending  its  energies  to  the  type  of  tasks  characteristic  of  a 
Kondratieff  downgrade,  the  federal  government  was  pursuing  a  fiscal 
policy  eminently  "sound"  in  the  old  sense.  It  reduced  taxation,  going  a 
considerable  way  beyond  merely  eliminating  the  excess  profits  tax;  it 
reduced  the  federal  debt  and  even  set  about  to  effect  some  retrenchments 
— Gladstone  himself  could  perhaps  have  acted  more  brilliantly,  but 
hardly  more  soberly.  Up  to  an  income  of  $100,000,  the  income  tax 
was  far  below  the  European  level.  Federal  expenditure,  which  in  1912 
-1913  had  been  724.5  millions,  moved,  it  is  true,  on  a  level  of  about 
3.7  billions  from  1925-1926  to  1929-19301  (including  debt  redemptions 
out  of  current  revenue  to  the  average  amount  of  over  %  billion  a  year2). 
But  under  general  conditions  so  exceptionally  favorable  this  was  not  a 
very  serious  matter.  Local  Total  Gross  Expenditure  increased  from 
4,593  millions  in  1923  to  6,720  millions  in  1929;  State  Total  Gross 
Expenditure,  from  1,208  to  1,943  millions.  But  both  states  and  local 
authorities  raised,  partly  from  a  lack  of  constitutional  powers,  partly 
from  choice,  the  money  they  spent  in  ways  which  (whatever  may  be 

1  Federal  Total  Ordinary  Receipts  (as  per  Treasury  Statements)  were:  1920,  6,694.6 
millions;  1921,  5,624.9;  1922,  4,109.1;  1923,  4,007.1;  1924,  4,012.0;  1925,  3,780.1;  1926, 
3,962.8;  1927,  4,129.4;  1928,  4,042.3;  1929,  4,033.3;  of  which  Tax  Collections  (National 
Industrial  Conference  Board,  Cost  of  Government,  1923-1934,  p.  18)  were:  1922,  3,487.0; 
1923,  3,032.0;  1924,  3,193.0;  1925,  2,966.0;  1926,  3,207.0;  1927,  3,337.0;  1928,  3,194.0;  1929, 
3,328.0.     Total  Gross  Expenditure  was  (Treasury  Statements)  6,482.1  millions  in  1920 
and  steadily  fell  to,  roughly,  3.5  billions  during  1924-1927,  after  which  it  rose  again  (1929, 
3,884.5;  it  was  5,153.6  in  1932). 

2  Those  redemptions  were  about  compensated  by  the  increase  in  state  and  municipal 
debt.     Total  public  debt  decreased  (slightly)  only  in  1923  and  for  the  rest  of  the  period 
continued  to  increase,  although  total  public  issues  outstanding,  minus  holdings  of  the 
United  States  government  and  government  trust  funds,  slightly  declined,  moving  on  a 
level  of  about  30  billions.     Therefore,  we  need  not  bother  about  the  question  how  net 
redemption  would  have  affected  money-market  situations  and,  through  them,  speculation 
and  business.     That  would  have  depended  on  the  sources  of  the  funds  by  which  redemption 
was  effected  and  on  the  behavior  of  the  households,  firms,  and  banks  that  held  the  bonds 
redeemed.     By  drawing  up  a  schema  of  possible  combinations,  the  reader  should  have  no 
difficulty  in  developing  a  complete  theory  of  the  effects  of  debt  redemption  and  in  realizing 
that  both  "deflationary"  and  "inflationary"  consequences  may  ensue  from  it. 


710  BUSINESS  CYCLES 

thought  about  them  from  other  standpoints)  did  not  substantially 
injure  the  economic  machine. 

Moreover,  the  government  promptly  abolished  most  of  the  wartime 
controls,  regulations,  and  organizations;  refrained  from  measures  involv- 
ing questions  of  social  and  economic  structure  at  home;  and  successfully 
kept  out  of  entanglements  abroad,  thereby  creating  the  atmosphere  con- 
genial to  private  business1  and  reducing  the  importance  to  the  American 
citizen  of  the  struggles,  sufferings,  and  upheavals  in  other  p$rts  of  the 
world  to  the  order  of  importance  of  a  football  match.  Economists  who 
are  passionately  determined  not  to  admit  that  policies  answering  to  their 
social  and  moral  vision,  particularly  fiscal  policies  of  anticapitalist 
tendency,  can  possibly  interfere  with  the  working  of  the  economic  system,2 
will  no  doubt  hold  that  there  was  mere  chance  coincidence  between  that 
sociopolitical  pattern  and  the  economic  results  achieved  in  this  country 
during  the  twenties,  and  between  the  different  setup  and  the  different 
results  in  England  or  Germany.  In  the  fulfillment  of  our  humble  task 
of  interpreting  a  given  course  of  historical  events  and  the  behavior  of 
given  time  series,  we  cannot,  however,  neglect  the  possible  inference  to 
the  contrary.  We  speak  of  possible  inference  only,  because  in  this  point 
our  argument  transcends  exact  proof,  as  any  argument  about  organic 
processes  occasionally  must,  and  because  so  many  imponderable  elements 
enter  which  must  be  a  matter  of  personal  judgment  and  (historical  and 
personal)  experience. 

2.  But  the  main  points  at  issue  with  reference  to  effects  of  taxation 
as  such — i.e.,  as  distinguished  from  those  effects  which  a  system  of 
taxation  may  have  if  it  is  or  is  felt  to  be  an  element  of  a  general  atmosphere 
of  hostility  to  capitalist  success3 — may  conveniently  be  mentioned  here 

1  It  has  been  pointed  out  to  the  writer  that  the  above  argument  reads  like  an  advocacy 
of  partisan  principles.     He  does  not  see  that.     Nor  does  he  see  how  he  could  guard  against 
this  danger,  short  of  inserting  explanatory  parentheses  to  an  impossible  extent.     If  the 
reader  misunderstands  that  passage,  he  will  misunderstand  so  much  in  this  book  that 
one  more  misunderstanding  hardly  matters.     Let  it  be  reemphasized,  however,  that  the 
above  (which,  besides,  contains  only  statements  of  obvious  fact  to  which  nothing  will  be 
added  except  an  inference  about  a  causal  connection)  neither  is  intended  to  convey,  nor  is 
as  a  matter  of  fact  sufficient  for,  an  appraisal  either  of  capitalism  or  of  the  particular 
line  of  action  taken  by  certain  political  groups  within  a  given  historic  situation. 

2  They  could,  however,  as  we  shall  not  tire  of  pointing  out  again  and  again,  admit  it 
without  prejudice  to  their  standpoint,  which  obviously  rests  on  extraeconomic  valuations. 
What  is  an  injury  to  an  organism  from  the  standpoint  of  a  doctor,  may  be  most  "desirable" 
from  plenty  of  other  standpoints. 

8  To  this  aspect  we  shall  have  to  return  in  the  last  chapter.  Also,  the  above  remarks 
cover  only  part  of  what  for  our  purpose  it  is  necessary  to  say  on  the  subject.  The  rest  has 
been  divided  up  and  inserted  in  various  places  in  such  a  manner  as  to  minimize  expenditure 
of  space.  It  includes  some  remarks  on  the  problem  how  the  use  made  of  the  sums  raised 
by  the  receiving  authority  influences  the  effects  of  the  "burden,"  on  the  view  that  taxation 
for  certain  purposes  is  "mere  transference,"  and  on  other  topics. 


1919-1929  711 

once  for  all.  Since  we  cannot  fully  go  into  the  matter,  we  will  avail 
ourselves  of  the  fact  that  there  is  (comparative)  agreement  about  the 
effects  of  indirect  taxes,  such  as  specific  taxes  on  the  quantity  produced 
or  sold  of  a  commodity.  This  agreement,  such  as  it  is,  we  owe  to  a  fairly 
well  elaborated  theory  which,  though  antiquated,  is  still  widely  accepted 
by  economists  and  has  recently  been  somewhat  improved  by  borrowings 
from  the  theories  of  imperfect  competition,  of  expectation,  and  so  on. 
Its  assumptions,  however,  limit  its  results  to  the  case  of  small  taxes  and /or 
of  individual  commodities  of  small  importance.  The  technical  reason 
for  this  has  an  important  counterpart  in  real  life:  wherever  taxes  are  so 
small  as  to  be  amenable  to  analytical  treatment  by  the  calculus,  they  are 
also  too  small  to  affect  the  fundamental  contours  of  economic  behavior  as 
reflected  in  the  budgets  of  firms  and  households  and,  hence,  to  interfere 
significantly  with  economic  processes  in  general  and  the  cyclical  process 
of  evolution  and  its  permanent  results  in  particular.  This  proposition 
may  be  generalized  to  cover  any  small  tax,  no  matter  whether  sectional, 
such  as  a  tax  on  beer  or  on  house  room,  or  general,  such  as  a  turnover  or  an 
income  tax,  and  extended  in  most  cases  to  any  tax  that  is  small  in  a  prac- 
tical— though  loose — sense  and  not  only  in  the  sense  of  the  calculus. 
Most  taxes — not  strictly  all — which  are  not  small  in  that  wider  sense,  on 
the  one  hand  cannot  be  handled  by  that  method — further  repercussions, 
more  fundamental  changes  in  the  economic  system,  reactions  from  and 
through  the  sphere  of  money  and  credit  must  be  then  taken  into  account 
— and  on  the  other  hand,  do  interfere  with  the  results  of  business  proc- 
esses, for  example,  with  the  steady  rise  in  the  standard  of  living  of  the 
masses  as  far  as  it  is  due  to  the  working  of  the  capitalist  machine.1 

This,  however,  marks  the  point  at  which  disagreement  begins.  The 
fiscal  problem  of  our  time  does  not  primarily  consist  in  the  amount  of 
revenue  required  by  the  modern  state,  but  in  the  fact  that,  owing  to  the 
moral  valuations  prevailing,  that  amount  must  also  be  raised  by  heavy 
taxes2  and,  moreover,  by  heavy  taxes  framed  not  only  without  a  view  to 

1  Within  limits,  other  methods  of  attaining  a  maximum  standard  for  the  masses  may  be 
effective  without  interfering  with  the  contribution  made  automatically  by  the  capitalist 
machine — the  machine  can  be  made  to  yield  more  than  it  would  of  itself  without  materially 
losing  in  efficiency.     Even  if  it  does  lose  in  efficiency,  the  advantage  to  the  labor  interest 
may  more  than  balance  the  loss  it  suffers,  especially  in  the  short  run,  but  also  in  the  long 
run.     All  that  the  proposition  of  our  text  amounts  to  is  that  any  such  advantage  is  gross 
only  and  must  be  subjected  to  scrutiny  as  to  concomitant  losses  to  be  deducted.     It 
stands  to  reason  that  the  advantages  will,  as  a  rule,  be  visible  and  immediate  and  the 
losses  more  difficult  to  see  and  more  remote  in  time. 

2  "Big"  revenue  need  not  imply  "heavy"  taxes.    A  transaction  tax,  for  instance, 
which  in  this  country  has  been  actually  proposed  in  the  McGroarty  bill,  would  go  far 
toward  raising  adequate  revenue  without  significantly  altering  the  conditions  of  economic 
progress.     According  to  Mr.  Golden weiser,  total  monetary  transactions  amounted  in  1929 
to  about  1,200  billions  (debits  to  individual  accounts  in  141  cities,  plus  an  estimated  addition 


712  BUSINESS  CYCLES 

minimum  disturbance  but  regardless  of  disturbance,  in  some  cases  even 
with  a  view  to  maximizing  it.  And  the  disagreement  that  is  relevant  to 
our  purpose  concerns  either  the  reality  of  the  effects  alluded  to  in  the  last 
sentence  of  the  preceding  paragraph  or  their  importance  for,  let  us  say 
a  potiorij  the  development  of  total  output.  We  will  confine  ourselves  to 
the  case  which  is  most  important  in  this  connection  and  consider  a  high 
and  highly  progressive  income  tax — by  which,  to  fix  ideas,  we  will  mean 
an  income  tax  which,  for  a  significant  number  of  taxpayers  in  the  higher 
and  highest  brackets,  surpasses  25  per  cent — that  so  defines  income  as  to 
include  savings  and  is  reinforced  by  a  significant  corporation  and  a  high 
or  highly  progressive  inheritance  tax. 

First,  there  are  what  we  may  term  mechanical  effects,  of  which  the 
most  important  is  the  effect  on  the  sum  total  of  private  savings  and 
accumulations.  Taxes  such  as  those  we  have  in  mind  may  enforce  dis- 
saving and  even  divestment,  but  will  in  general  be  partly  paid  from 
revenue  that,  in  turn,  would  otherwise  be  partly  saved.  An  obvious 
argument  from  general  principles  yields  the  result  that,  as  a  rule,  this 
again  will  be  partly  made  up  for  by  additional  saving  by  the  same  people 
or  by  those  who  are  the  ultimate  recipients  of  the  sums  levied.  But  as 
far  as  the  writer  knows,  nobody  has  so  far  doubted  that  the  net  effect  of 
high  taxes  on  the  higher  incomes  will  be  a  decrease  of  the  national  total 
of  savings  as  compared  with  what  it  otherwise  would  be.  As  far  as  this 
goes,  therefore,  our  opinion  on  how  such  taxes  will  affect  "progress" 
and  "industrial  efficiency"  depends  on  where  we  stand  in  the  controversy 
about  the  importance  and  the  modus  operandi  of  private  saving,  which 
have  been  fully  discussed  before.1 


for  checks  drawn  on  banks  outside  those  cities  and  payments  in  cash).  %  per  cent  of  this 
sum  or,  if  certain  classes  of  payments  were  to  be  excluded,  1  per  cent  of  half  of  it  would  be  a 
small  tax  in  the  sense  that  it  would  nowhere  exert  perceptible  pressure.  But  it  would 
yield — in  a  year  like  1929 — 6  billions. 

Big  taxes  interfere  with  the  economic  process  to  greatly  varying  degrees.  Big  taxes 
on  individual  spots,  such  as  a  heavy  tax  on  alcoholic  drinks,  does  interfere  significantly 
with  the  particular  sector  on  which  it  impinges  but  not  with  the  system  as  a  whole.  Among 
"general"  big  taxes,  a  tax  on  the  return  to  or  value  of  natural  agents  would  be  a  good 
instance  of  a  limiting  case  in  which  no  long-run  effects  are  to  be  expected  if  it  were  not  for 
the  facts  that  improvements  are  not  easy  to  distinguish  from  "pure  rent"  and  that  the 
latter,  as  we  have  seen,  for  example,  in  the  case  of  the  apartment-house  industry  of  Berlin, 
steps  into  the  place  of  and  acts  as  a  substitute  for  profit.  But  all  that  matters  here  is  the 
technical  possibility  of  raising  "big"  revenue  with  small  disturbance  if  that  were  intended — 
and  small  revenue  with  big  disturbance. 

1  But  it  should  be  observed  that  many  arguments  turn,  not  on  saving  in  our  sense, 
but  on  underspending.  Taxes  on  idle  funds  may  have  some  stimulating  short-run  effects 
if  conceived  as  temporary  measures.  This  point,  too,  has  been  fully  discussed  in  Chap. 
XI.  One  more  aspect  of  saving  and  accumulation,  which  is  important  for  an  appraisal  of 
short-run  effects,  will  be  added  in  the  next  chapter, 


1919-1929  713 

Second,  there  are  the  nonmechanical  effects,  i.e.,  the  effects  through 
motives  and  attitudes.  It  should  be  obvious  that  any  tax  on  net  earnings 
will  tend  to  shift  the  balance  of  choice  between  "to  do  or  not  to  do"  a  given 
thing.  If  a  prospective  net  gain  of  a  million  is  just  sufficient  to  over- 
balance risks  and  other  disutilities,  then  that  prospective  million  minus 
a  tax  will  not  be  so,  and  this  is  as  true  of  a  single  transaction  as  it  is  of 
series  of  transactions  and  of  the  expansion  of  an  old  or  the  foundation 
of  a  new  firm.  It  should  be  equally  obvious  that  business  management 
and  enterprise,  being  undertaken  within  an  institutional  framework  of 
aims,  ambitions,  and  social  values  fashioned  to  its  logic,  will  for  its 
maintenance  depend,  at  least  in  the  long  run,  on  the  actual  delivery,  in 
case  of  success,  of  the  prizes  which  that  scheme  of  life  holds  out,  and  that, 
therefore,  taxes  beyond  a  percentage  that  greatly  varies  as  to  time  and 
place1  must  blunt  the  profit  motive  and,  especially,  the  motive  typical  of 
both  feudal  and  bourgeois  society,  that  of  founding  a  family  position. 
As  to  the  profit  motive  in  general,  it  must  be  borne  in  mind  that  a  policy 
of  taxing  away  gains  evidently  above  what  would  be  necessary  to  call 
forth  the  efforts  of  their  individual  recipients  and  of  taxing  but  moderately 
what  the  community  considers  "adequate"  returns,  if  it  is  not  to  affect 
the  total  amount  of  effort,  would  really  have  to  be  accompanied  by  an 
increase  in  the  sum  total  of  managerial  and  entrepreneurial  income, 
because  the  presence  of  conspicuously  high  and  even  fantastic  individual 
prizes  is,  as  everyone  knows,  much  more  stimulating  than  the  same  sum 
would  be  if  more  equally  distributed  among  businessmen.  As  to  that 
special  form  of  the  profit  motive  which  is  embodied  in  the  term  family 
position,  and  is  largely  eliminated  by  inheritance  taxes  of  the  modern 
type,  it  is  as  reasonable  to  hope  that  high  inheritance  taxes,  being  taxes 
on  "static"  wealth,  will  not  affect  industrial  "progress,"  i.e.,  the  creation 
of  new  wealth,  as  it  would  be  to  hope  that  a  prohibitive  railroad  fare  will 
not  affect  traffic  if  passengers  be  allowed  to  board  the  trains  free  of  charge 
and  the  fare  be  collected  from  them  after  they  have  taken  their  seats. 

The  reader  has  a  right  to  object  to  the  triviality  of  these  consider- 
ations. Yet  it  is  a  fact  that  the  reality  of  effects  through  motives  and 

1  Moderate  taxation,  i.e.,  taxation  which,  while  making  it  more  difficult,  yet  does  not 
make  it  too  hard  to  attain  a  given  economic  position,  may  even  act  as  a  stimulus.  But 
however  difficult  it  may  be  to  determine  the  interval  for  which  that  is  so,  it  is  perfectly 
clear  that  since  the  war  taxation  in  the  higher  brackets  goes  much  beyond  it. — High  taxa- 
tion, for  example  in  a  national  emergency,  as  long  as  it  is  considered  to  be  temporary,  may 
have  no  effect  on  motive  or  even  an  effect  that  is  stimulating.  What  taxation  is  "high"  and 
what  "moderate"  also  depends  on  the  prevailing  margins  of  profits.  American  taxation 
even  from  1924  to  1031  might  have  been  high  in  our  sense  but  for  the  ease  with  which  the 
businessman  rode  to  success.  Finally,  much  depends  on  the  reaction  of  the  monetary 
system,  for  example,  on  whether  or  not  taxpayers  are  willing  and  able  to  borrow  the 
amounts  they  have  to  pay. 


714  BUSINESS  CYCLES 

attitudes  is  still  more  frequently  denied  than  the  bearing  of  accumulation 
on  industrial  efficiency.1  But  there  is  at  least  an  argumentum  ad  hominem 
to  be  addressed  to  any  economist  who  uses  the  profit  motive  in  his  analysis 
at  all.2  For  net  revenue  would  have  to  be  a  matter  of  complete  indiffer- 
ence if  rates  such  as  we  now  discuss  had  no  effect  on  the  working  of  the 
capitalist  machine.  If,  moreover,  that  economist  teaches,  as  he  is  likely 
to  do,  the  omnipotence  of  comparatively  small  changes  in  the  rate  of 
interest  and  the  effectiveness  in  stimulating  industry  of  protection,  sub- 
sidies, and  *  deflationary"  increases  of  prices,  and  still  denies  that  high 
taxation  has  any  effect  whatever  on  the  scale  of  output  or  "progress/* 
he  comes  pretty  near  to  contradicting  himself,  unless  he  holds  that  that 
effectiveness  completely  vanishes  at  the  watershed  between  profit  and 
loss. 

3.  The  case  of  Germany  is  less  easy  to  describe  than  that  of  the 
United  States.  Refraining  from  a  repetition  of  facts  that  are  but  too 
well  known  (although  some  are  perhaps  already  forgotten)  and  taking 
our  stand  on  the  year  1925,  we  may  thus  characterize  the  situation. 
Menace  of  social  breakdown  had  been  warded  off  by  the  determined 
action  of  the  Social  Democratic  party  in  1918  and  1919,  which  thereupon 
found  itself  in  the  impossible  position  of  being  (whether  in  office  or  not) 
the  dominating  political  power  in  a  capitalist  society  which  it  had  saved 
when  all  that  society's  political  organs  were  utterly  shattered,  but  which 
it  was  by  virtue  of  its  principles  unable  to  run  according  to  capitalist 
logic.3  The  result  was  that  economic  policy  could  not  be  rationalized 

1  Many  economists  would,  however,  argue  that  the  effect  of  taxation  on  the  motives 
or  the  behavior  of  businessmen,  whatever  it  may  be,  can  be  neglected  in  an  analysis  of 
variation  in  output  and  employment,  because  these  motives  do  not  matter  anyhow,  the 
businessman's  hunt  for  profit  having  nothing  to  do  with  either  output  or  employment 
which  result  from  the  economic  process  except  that  it  sometimes  interferes  with  them.     A 
view  of  this  type  is  largely  a  vision  or  impression — easily  contracted  in  this  newspaper 
world  of  ours,  in  which  words  are  in  the  saddle — about  which  it  is  difficult  to  argue  except 
by  discussing  the  individual  facts  that  in  each  case  would  be  produced  in  response  to  a 
request  for  substantiation.     In  a  general  way,  however,  our  analysis  as  a  whole  supplies 
the  elements  for  criticism  of  this  vision  as  well  as  the  elements  of  its  sociology. 

2  Marshall's  great  shade  cannot  be  conjured  up  to  testify  against  that.     For  though 
he  seems  to  have  held  the  opinion  that  a  liberal  dose  of  direct  taxation  would  not  reduce 
industrial  efficiency,  the  rates  he  visualized  were,  except  for  temporary  emergencies,  so 
low  as  to  make  his  argument  quite  compatible  with  ours.     Nor  can  it  be  urged  that  we  are 
using  a  rational  schema  of  behavior  that  does  not  take  account  of  fixed  habits.     We  do  take 
account  of  them.     But  this  particular  application  of  economic  rationality  is  surely  not 
unrealistic  or  farfetched.     And  those  habits  are  likely  to  give  way  under  the  influence  of 
unfavorable  experience  and  pessimistic  anticipation,  as  do  even  the  habits  of  a  puppy, 
which  does  not  indefinitely  jump  for  a  sausage  that  is  pulled  out  of  his  reach  each  time. 
Were  it  not  for  the  reluctance  to  admit  unpalatable  facts,  there  would  be  little  disagreement 
about  this. 

8  Socialists  went  to  lengths,  however,  which  amounted  to  an  opus  super  erogationis. 
The  Sozialisierungsgesetz  (1019)  meant  the  shelving  of  socialization.     But  more  was  to 


1919-1929  715 

either  in  the  socialist  or  in  the  capitalist  sense:  a  deadening  laborism 
threatened  everybody  and  satisfied  nobody.1  Menace  of  national 
destruction  had  been  warded  off  by  the  guarantees  against  invasion 
included  in  the  Dawes  plan.  Menace  of  economic  deadlock  had  been 
warded  off  by  those  credits  which  the  world  pumped  into  Germany,  obvi- 
ously believing  that  credits  and  democracy  were  all  that  a  nation  could 
possibly  need.  Menace,  finally,  of  irreparable  moral  disorganization 
through  inflation  had  been  warded  off  by  an  energetic,  if  clumsy,  bal- 
ancing of  the  budget,  consequent  upon  the  construction  of  a  rather  rigid 
gold  standard,  both  the  Dawes  loan  and  the  successful  bluff  of  the 
Rentenmark  being  minor,  though  still  important,  safeguards  or  pieces  of 
technique. 

The  complete  loss  of  foreign  investments  and  of  industrial  and  com- 
mercial positions  abroad  would  in  any  case  have  been  sufficient  to 
unbalance  an  economy  so  largely  based  on  them  as  the  German  was. 
Moreover,  Germany's  industrial  organism  emerged  from  the  war  with  an 
antiquated  productive  apparatus,  which  then  suffered  further  injury 
and,  in  part,  further  distortion  because  of  that  kind  of  irrational  invest- 
ment the  sole  motive  of  which  was  to  take  shelter  from  inflation — a  good 
illustration  of  the  popular  belief  that  if  only  there  be  expenditure  it  does 
not  matter  what  direction  it  takes.  A  crisis  of  adjustment  both  of 
structures  and  of  values  was  bound  to  set  in  to  clear  the  ground  as  soon 
as  the  support  of  progressive  inflation  was  withdrawn,  but  much  mal- 
adjustment then  in  existence  was,  as  it  always  must  be,  very  slow  in 
dying.  To  the  end  of  the  period  some  of  the  growths  of  inflation  con- 
tinued their  sickly  life,  constituting  weak  points,  sources  of  chronic 

follow.  The  minister  Hilferding,  much  too  good  an  economist  not  to  see  what  was  wrong 
and  much  too  good  a  Marxist  not  to  realize  that  there  are  situations  in  which  anticapitalist 
policy  is  in  the  end  antisocialist,  actually  went  so  far  as  to  attempt  a  very  "capitalistic" 
fiscal  reform.  In  November  1927  the  leading  socialist  newspaper  advocated  reduction 
of  the  income  tax  (in  all  brackets).  Nothing  came  of  it.  Difficulties  for  the  socialists 
were  greatly  increased  by  the  fact  that  their  political  allies,  the  centrist  party,  though  less 
radical  on  principle,  were  much  less  amenable  to  economic  reasoning  than  were  the  socialists 
themselves. 

1  The  picture  cannot  be  developed  here,  but  two  features  must  be  mentioned  in  passing. 
First,  the  trade  unions  secured  two  points  of  their  program — the  eight-hour  day  and  unem- 
ployment insurance.  These  measures  were,  in  fact,  overdue.  But  beyond  that  a  huge 
legislative  and  administrative  apparatus  was  built  up  by  the  competitive  exertions  of  the 
two  ruling  parties  for  the  purpose  of  serving  the  immediate  interests  of  labor  (the  Arbeits- 
rechi),  which,  whatever  its  merits,  is  relevant  to  our  subject  by  virtue  of  the  economic  waste 
and  friction  it  created,  and  particularly  by  virtue  of  the  efficiency  with  which  it  "skimmed," 
as  the  Minister  Braun — a  centrist — put  it,  the  results  of  every  upturn  in  business  for  the 
immediate  benefit  of  labor,  i.e.,  applied  potential  capital  to  consumptive  purposes.  Second, 
the  socialists  had  to  accept  a  no  less  wasteful  agrarian  policy,  including  subsidies  to  eastern 
agriculture,  which  weighed  heavily  on  the  economic  process,  yet  entirely  failed  to  reconcile. 


716  BUSINESS  CYCLES 

trouble,  and  one  of  the  difficulties  in  Germany's  economic  relations  with 
other  countries. 

The  weakness  of  her  international  position  made  her  industries 
particularly  anxious  to  enter  into  agreements  with  foreign  competitors  in 
order  to  eliminate  their  pressure  for  hostile  measures.  Some  temporary 
success  was  achieved1  but,  as  far  as  it  was,  it  carried  all  the  disadvantages 
incident  to  "stabilization"  in  creating  rigidities  and  cramped  situations. 
From  1925  to  1928,  however,  the  surface  of  things  was  smoothed  by 
those  foreign  credits,  by  comparatively  moderate  impediments  to  trade, 
and  by  comparative  order  in  monetary  systems  all  over  the  world,  as 
pointed  out  above.  Even  then  the  risk  of  industrial  investment  remained 
in  Germany  much  greater  than  it  was  in  England  or  in  the  United  States, 
and  much  waste  of  effort  and  resources  ensued  from  the  necessity  of 
shifting  locations  of  industries  according  to  political  fears,  and  of  remodel- 
ing equipment  in  response  to  foreign  measures — an  element  never  fully 
appreciated  in  diagnoses  of  the  German  situation.  But  when  the  flow  of 
foreign  capital  ebbed  and  in  1929  finally  ceased,  and  Germany's  balance 
of  commodity  trade  promptly  and  easily  (without  any  "stickiness") 
swung  in  her  favor  in  consequence2 — a  most  striking  verification  of  classic 
theory  of  foreign  trade,  as  was  also  a  temporary  tendency  in  this  direction 
in  1926 — that  smooth  surface  went  to  pieces.  German  reparation 
exports  had  suddenly  become  a  reality.  They  were  one  of  the  influences 
that  shaped  the  international  commercial  situation  on  the  threshold  of 
the  world  crisis,  helped  to  induce  the  vicious  circle  of  mutual  restrictions 
(quota  and  so  on),  and  effectively  brought  out  the  absurdity  of  insisting 
on  payments  which  the  creditor  countries  at  the  same  time  stoutly  refused 
to  accept. 

As  long  as  they  continued  to  flow,  however,  foreign  credits  smoothed 
not  only  the  difficulties  incident  to  transfers  on  reparation  account  but 
other  difficulties  as  well.  In  order  to  understand  these  it  is  necessary 
to  recall  what  has  been  said  about  the  social  atmosphere  and  the  political 

1  By  the  end  of  1926  a  considerable  list  of  international  cartels  was  in  force,  such  as  the 
potash  convention  of  Lugano,  the  European  union  of  bottle  manufacturers,  other  European 
unions  of  enamel  works  and  wood-screw  producers,  the  international  bulb  syndicate,  the 
international  rail  cartel,  the  cartel  of  German  and  Czechoslovakian  iron -pipe  producers, 
the  German- Belgian  wire  cartel.     International  organizations  for  iron,    copper,  benzol, 
and  other  articles  were  at  the  same  time  under  consideration,  and  some  of  them  actually 
matured.     After   1927,   when  French  interests  became  increasingly    restive    under  the 
impact  of  German  exports,  machinery  was  provided  for  producers  of  both  nations  to  meet 
and  to  find  themselves  a  modus  vivendi  (quota)  by  mutual  agreement,  to  be  ratified  by  the 
respective  governments.     For  some  time  this,  too,  seemed  to  work  with  comparative 
success. 

2  Export  surplus:  minus  1,725  million  marks  in  1928;  plus  36  in  1929;  1,642  in  1930; 
2,872  in  1931.     After  that  it  began  to  fall  owing  to  the  universal  "mcapsulation"  incident 
to  the  world  crisis. 


1919-1929  717 

structure  of  that  epoch,  and  to  look  at  the  fiscal  policy  that  resulted 
therefrom.  The  first  federal  budget  after  inflation,  though  it  intensified 
the  "crisis  of  stabilization"  by  cutting  deeply  into  the  working  capital 
of  industry,  was  a  signal  success,  yielding  as  it  did  a  surplus  of  about  one 
billion  marks.  In  1925  the  total  expenditure  of  the  federal  government, 
the  states,  and  the  municipalities,  including  social  insurance  and  repara- 
tion payments,  amounted  to  17.3  billion  marks  or  (as  officially  estimated) 
about  31.9  per  cent  of  the  national  income,  as  compared  to  8.4  billions 
or  18.9  per  cent  (postwar  territory)  in  1913. l  But  Germany's  new  political 
structure  was  unable  to  withstand  popular  demands  and  to  plan  rationally 
for  the  future.  As  soon  as  the  immediate  emergency  was  overcome,  and 
the  unwieldy  mass  of  fiscal  legislation  systematized  and  somewhat 
adjusted  by  the  "reform"  of  1925,  that  fact  asserted  itself  exactly  as  it 
did  in  France  after  1928.  Notwithstanding  the  friendly  help  which  the 
Reparation  Agent  extended  to  the  government  by  means  of  unfriendly 
notes  and  reports,  expenditure  increased  by  leaps  and  bounds.  It 
reached  the  figure  of  23.3  billions  in  1927  and  from  that  year  on  entailed  a 
deficit,  from  1929  even  an  embarrassing  shortage  of  cash,  although 
public  revenue  steadily  increased  through  1929,  in  spite  of  drastic  reduc- 
tions in  the  turnover  tax  that  had  wrought  the  miracle  of  1925.  Total 
indebtedness  of  public  bodies  rose,  apart  from  the  partial  revaluation  of 
prewar  and  war  debts,  to  7.7  billions  in  1927—1928  and  by  another 
6  billions  net2  till  the  end  of  1929. 

Analysis  of  the  expenditure  thus  financed  undoubtedly  reveals 
admirable  cultural  and  social  achievement,  eminently  productive  of 
economic  and  supereconomic  values,  in  comparison  with  which  the  costs 
might  even  be  called  moderate.  In  particular,  there  is  much  to  be  set, 
in  terms  of  beauty  as  well  as  in  terms  of  welfare,  against  the  desperate 
financial  position  into  which  the  big  cities  maneuvered  themselves. 
But  neither  that  cultural  aspect  nor  the  various  deficits  as  such  are 
pertinent  to  our  subject.  The  important  thing  is  the  unavoidable  infer- 
ence that  we  have  here  a  case  of  an  excess  of  consumption  by  public 
bodies,  inducing  excesses  all  over  the  economic  system,  withdrawing 
capital  from  industry,  or  preventing  its  being  built  up — directly  by 

1The  figures  refer  to  what  is,  in  Germany,  technically  described  as  Finanzbedarf. 
They  include  items  such  as  interest  on  the  public  debt,  which  it  is  usual  to  class  as  "mere 
transferences."  Taxes  plus  contributions  to  social  insurance  amounted  respectively  to 
18.3  billions  or  24.5  per  cent  in  1925  and  5.4  billions  or  11.5  per  cent  in  1913.  They 
increased  to  19.6  billions  or  27.8  per  cent  in  1929.  See  for  all  figures,  for  example,  Statis- 
tisches  Reichsamt,  Finanzen  und  Steuern  in  In-und  Ausland  1930,  p.  548.  The  figures  for 
national  income  are  highly  controversial  (see  Sec.  E). 

2  Increase  from  Apr.  1,  1928,  to  Dec.  31,  1929,  was  6.56  billion  marks,  but  redemption  of 
prewar  and  war  debts  amounted  to  about  560  millions.  Those  debts  in  foreign  currency 
or  in  units  guaranteed  against  depreciation  (Festwert-und  Valutaschulden),  which  were 
contracted  during  the  war  and  inflation,  are  excluded. 


718  BUSINESS  CYCLES 

taxation,1  indirectly  by  the  ensuing  rise  in  costs — and  another  illustration 
of  the  Doctrine  of  Spending.  While  it  lasted,  this  process  of  impoverish- 
ment produced  what  many  Germans  at  the  time  quite  well  characterized 
as  a  "prosperity  of  consumption"  (Konsum-Konjunldur)  which,  con- 
trasting so  strikingly  with  underlying  difficulties,  superimposed  itself 
on  the  ordinary  run  of  cycles  and  produced  figures  of  output  that  almost 
incurred  the  Dawes  plan  penalties.  No  doubt  many  fellow  economists 
will  call  that  the  most  normal  state  of  things  imaginable  and  deny  with 
due  epithets  any  suggestion  to  the  effect  that  it  would,  in  any  case,  have 
had  to  end  either  in  a  breakdown  or  in  "inflation"  followed  by  a  still 
more  vehement  breakdown. 

4.  But  the  particular  manner  in  which  the  breakdown  actually  came, 
as  well  as  the  effect  of  that  policy  before  it  came,  was  determined  by  the 
flow  of  foreign  (and  expatriated  German2)  balances.  Data  are  not 
entirely  reliable,  but  the  main  facts  about  this  capital  import  stand  out 
clearly  enough.3  "Restriction"  having  been  decreed,  i.e.,  "inflation" 
having  been  definitively  renounced  by  the  Reichsbank  on  Apr.  7,  1924, 
both  demand  for  foreign  credits  and  readiness  to  grant  them  manifested 
themselves  almost  immediately.  Although  prevented,  until  the  autumn 
of  1925,  by  the  embargo  on  foreign  lending  from  offering  German  secur- 
ities for  public  subscription,  English  banks  and  bankers  displayed  willing- 
ness to  resume  short-term  relations  with  former  German  customers  as 
soon  as — and  before — the  Dawes  plan  was  carried,  and  even  a  ten-year 
private  loan  was  negotiated  in  August  1924  (Prudential  Insurance — 
North  German  Lloyd).  American  capital  took  nine-tenths  of  the  loan 
to  Krupp  soon  after  the  flotation  of  the  reparation  loan.  These  trans- 

1  The  political  structure  also  proved  to  be  unequal  to  its  task  in  that  it  reduced  unpop- 
ular taxes  while  keeping  up  those  that  injured  the  economic  engine.     For  1929  the  total 
load  of  taxation  resting  on  industrial  net  returns,  including  everything,  was  about  90  per 
cent  in  the  highest  brackets,  in  many  cases  more  than  that.     See  No.  4  of  the  monographs 
(Einzelschriften)  published  by  the  Statistische  Reichsamt.     It  does  not  follow,  of  course, 
that  without  those  taxes  net  returns  would  have  equaled  the  figures  from  which  the  per- 
centages were  calculated. 

2  Both  during  and  after  inflation,  the  flight  of  "capital"  from  Germany  must  have  been 
considerable,  and  part  of  these  balances  probably  returned  under  foreign  flags.     Figures 
are  not  available,  but  this  matters  little  for  our  argument,  since  expatriated  capital  pre- 
sumably behaved  much  as  did  bona  fide  foreign  funds. 

3  An  official  survey  of  foreign  indebtedness  was  first  made  in  1926.     After  the  break- 
down, however,  and  when  the  "standstill  agreements"  were  being  negotiated,  a  more 
comprehensive  investigation — as  of  July  28,  1931 — was  made,  and  some  further  data  were 
added  later.     The  international  commissions  appointed  in  pursuance  of  the  recommen- 
dations of  the  London  conference  (July  21  to  July  23,  1931)  issued  two  reports,  known 
respectively  as  the  Lay  ton  and  the  Beneduce  reports  (August  and  December  1931),  which 
presented  most  of  the  available  material.     Among  the  literature  on  this  subject  the  reports 
of  a  committee  of  the  Verein  fUr  Sozialpolitik  (Die  Auslands-Kredite  1928,  ed.  W.  Lotz, 
Schriften  vol.  174,  Chap.  Ill)  should  be  particularly  mentioned. 


1919-1929  719 

actions  ushered  in  a  long  series  of  similar  ones :  states,  provinces,  munici- 
palities, semipublic  corporations,  especially  light  and  power  concerns, 
public  and  semipublic  credit  institutions,  religious  bodies  of  all  types, 
as  well  as  banks  and  industrial  companies  all  rushed,  in  spite  of  the 
"suasion"  and  more  drastic  measures  of  the  Reichsbank,  for  the  oppor- 
tunity of  borrowing  long  in  terms  of  foreign  currency  at  what,  counting 
everything,  on  the  average  came  to  roughly  9  per  cent.1 

They  did  more  than  that,  however.  Partly  because  of  that  official 
resistance  to  the  rising  tide  of  foreign  indebtedness,  partly  for  obvious 
financial  reasons,  borrowers,  especially  concerns  and  banks,  at  the  same 
time  borrowed  short  in  whatever  happened  to  be  the  most  accessible  of 
the  foreign  money  markets.  These  short  loans  were  cheap  only  if 
granted  in  foreign  currency,  but  German  debtors  also  owed,  by  the  end  of 
July  1931,  about  4  billion  marks  to  foreign,  mostly  nonbanking,  creditors,2 
not  all  of  whom  were  bona  fide  foreigners.  The  inflow  of  these  various 
funds  was  a  function,  not  only  of  business,  but  also  of  political  situations 
and  fluctuated  considerably.  But  it  persisted  almost  to  the  end  of  1930. 
By  then  the  long-term  foreign  debt  amounted,  according  to  official 
estimate,  to  the  equivalent  of  9.2  billion  marks,  and  the  short-term 
foreign  debt,  after  some  repayments  in  consequence  of  the  scare  of 
1929,  to  about  14.9  billions.3  Other  foreign  investments  were  estimated 
at  6  billions.  This  total  of,  roughly,  30  billions  provided  the  exchange, 
first  of  all,  for  the  transfers  on  reparation  account  equivalent  to  10.3,  and 
for  interest  payments  which  amounted  to  2.5  billion  marks.  It  also 
financed  the  building  up  of  foreign  balances  and  investments — 9.7  billions 
and  possibly  more — and  the  net  increase  in  Germany's  gold  reserve 
(deducting  the  reduction  in  foreign  exchange  held  by  the  Reichsbank) 
which  was  2.1  billions.  Finally,  it  also  covered  the  deficits  in  her  com- 
modity balance  of  trade,  which  during  those  seven  years  added  up, 
after  deduction  of  3  billions  for  "services  rendered,"  to  3.3  billions. 
This  accounts  for  27.9  out  of  those  30  billions.  Various  hypotheses 
suggest  themselves  with  regard  to  the  difference.  It  is,  however,  not 

1  The  high  cost  to  borrowers  reflects,  of  course,  primarily  a  large  premium  for  risk. 
But  it  must  also  be  remembered  that  the  cost  of  most  of  these  transactions  was  very  high. 
Most  of  the  securities  placed  in  America,  for  instance — representing,  roughly,  two-thirds 
of  the  total  amount  of  the  long-term  foreign  debt — had  to  be  laboriously  sold  all  over  the 
country  to  people  of  moderate  means  since,  owing  to  the  progressiveness  of  the  American 
income  tax  and  to  the  tax  privileges  attaching  to  the  Liberty  loans,  the  higher  return  from 
German  bonds  was  not  so  attractive  to  large  investors  as  one  might  think. 

2  English  banks  had,  since  1927,  also  granted  some  credits  in  marks  to  German  banks. 
These  were  not  short,  however. 

8  The  official  survey  mentioned  in  a  previous  note  puts  them  at  12  billions,  inclusive 
of  the  4  billions  that  were  mark  credits.  But  that  figure  is  of  end  of  July  1931;  2.9  billions 
had  been  withdrawn  in  the  preceding  seven  months. 


720  BUSINESS  CYCLES 

greater  than  we  should  expect  it  to  be  with  data  of  this  nature,  which 
also  leave  many  other  points  in  doubt. 

Thus,  the  excess  of  imports  over  exports  absorbed  only  11  per  cent  of 
the  total  inflow  of  monetary  "capital."  Two-thirds  of  it — roughly,  20 
billions — was  used  up  by  reparation  payments  and  Germany's  foreign 
investment,  including  balances  with  foreign  banks.  It  is  not  super- 
fluous, however,  to  emphasize  that  the  effects  of  this  are  inadequately 
described  by  saying  that  reparations  and  foreign  investments  were 
"paid  out  of  foreign  loans."  This  applies  strictly  only  to  a  small  part 
of  that  sum,  at  most  2  or  3  billions,  which  were  directly  borrowed  for 
and  applied  to  those  purposes:  only  to  this  extent  the  economic  life  of 
Germany  was  for  the  time  being,  in  fact,  relieved  from  all  further  effects, 
exactly  as  if  someone  else  had  undertaken  to  carry  those  burdens  for  her. 
The  modus  operandi  of  all  the  rest  was  more  complicated.  The  sums 
required  were  actually  raised  by  taxation  or  paid  by  the  individuals  or 
firms  who  wished  to  invest  abroad,  and  all  the  foreign  credits  did  was  to 
provide  the  exchange  with  which  to  transfer  them.  Schematically  we 
can  represent  this  process  by  assuming  that  taxpayers  and  investors 
bought  the  foreign  exchange  which  borrowers  had  acquired,  handing  their 
marks  over  to  them  and  the  foreign  exchange  to  foreign  governments  or 
sellers  of  assets.  As  thus  financed,  reparations  and  foreign  investments 
neither  increased  nor  reduced  available  funds  in  Germany.  The  foreign 
credits  prevented  temporarily  all  those  adjustments  of  incomes,  prices, 
and  the  balance  of  commodity  trade  which  otherwise  would  have  resulted 
from  both.  But  they  prevented  them  by  a  route  that  was  different  from 
and  much  rougher  than  that  of  direct  borrowing  for  reparations  and 
investments.  This  particular  method  was  rendered  possible  by  the 
fact  that,  barring  import  surpluses  and  interest  payments,  German  busi- 
ness had  no  use  for  foreign  exchange  but  needed  marks  for  expenditure  at 
home.  This — the  salient — point  stands  out  still  more  clearly  with 
respect  to  that  amount  by  which  foreign  credits  surpassed  reparation, 
foreign  investment,  import  surplus,  and  interest  requirements,  and  which 
— partly  reflected  in  the  increase  of  the  Reichsbank's  gold  holdings — of 
course,  swelled  deposits. 

The  effects  are  clear:  not  only  were  adjustments  to  the  facts  of  the 
situation  prevented,  but  the  pulse  of  German  business  became  dependent 
on  the  rate  of  flow  of  foreign  funds ;  with  foreign  banks  indirectly  financing 
a  considerable  part  of  investment  and  current  operations  in  Germany,  the 
policy  of  her  central  bank  was  checkmated;  the  Konsum-Konjunktur 
mentioned  before  was  powerfully  propelled;  and  of  course,  a  financial 
situation  was  created  that  was  in  constant  danger  of  collapse  on  com- 
paratively small  provocation,1  especially,  as  it  was  hardly  avoidable  under 

1  Many  firms  carrying  on  purely  domestic  business  but  having  acquired  the  habit  of 
financing  themselves  by  short  foreign  credits  came,  however  prosperous  their  income 


1919-1929 

the  circumstances  that  both  German  and  foreign  banks  should,  directly 
and  indirectly,  knowingly  and  unwittingly,  finance  long-term  industrial 
commitments  by  short  foreign  funds.  Thus,  part  of  the  foreign  credits 
effected  precisely  what  an  issue  of  greenbacks  might  have  done:  in  a 
sense,  it  camouflaged  "inflation"  by  producing  its  results  under  the  sur- 
face of  an  apparently  very  "sound"  monetary  system. 

But  still  more  interesting  than  the  effects  are  the  causes  of  this  practice 
of  financing  Germany's  domestic  business  in  this  way.  Concerns  bor- 
rowed to  such  an  extent  because  taxation  absorbed  what  otherwise  would 
have  become  fixed  and  working  capital.  They  borrowed  abroad  because 
the  same  taxation — together  with  the  preceding  inflation — had  reduced, 
and  kept  on  reducing,  the  means  of  savers,  because  public  and  semipublic 
expenditure  absorbed  a  large  amount  of  the  lending  ("creating")  power 
of  German  financial  institutions  and  because  domestic  credit  was,  in 
consequence,  both  scarce  and  dear.1  The  responsibility  of  taxation  and 
of  public  expenditure  for  the  short  credit  situation,  in  particular,  is 
obvious.  Assuming  that,  at  the  end  of  1928,  foreign  short  credits 
amounted  to  13  or  14  billions — an  estimate  that  is  arrived  at  by  adding 
50  per  cent  to  the  figure  given  in  the  Layton  report — and  deducting 
(without  increasing  the  figure,  though  this  would  almost  certainly  be 
justified)  German  balances  abroad,  4.5  billions,  and  credits  for  the  current 
financing  of  German  foreign  transactions,  which  were  "revolving"  and 
not  dangerous — this  item,  being  very  doubtful  and  to  some  extent  also 
overlapping  with  the  first,  we  will  put  at  not  more  than  3  billions — we 
may  perhaps  estimate  at  a  figure  of  the  order  of  magnitude  of  6  billions 
the  credits  that  eventually  created  trouble.  How  much  of  these  does 
the  reader  think  would  have  flowed  in  at  all  if  total  public  (not  only 
federal)  expenditure  had  been  kept  at  the  1925  level,  if  the  surplus  of  that 
year  had  been  lent  in  the  open  market,  if  income  and  corporation  taxes 
had  been  reduced,  and  if  interest  rates  and  prices  had  behaved  as  in  that 
case  they  assuredly  would  have?  It  should  be  added  that,  though  the 
"consumers'  prosperity"  would  then,  no  doubt,  have  failed  to  come 


statement,  in  danger  of  bankruptcy  whenever  a  cloud  darkened  the  political  sky.  Since 
short  credits  will  always  take  to  flight  in  such  cases,  there  is  no  need  of  attributing  such 
phenomena  to  foreign,  e.g.,  French,  ill  will.  But  the  logic  of  the  situation  itself  created  a 
sensitiveness  of  large  sectors  of  business  to  erratic  shocks,  which  was  altogether  abnormal 
and  cannot  be  explained  on  any  of  those  principles  that  we  ordinarily  use  in  analyzing 
economic  fluctuations  and  "spirals." 

1  It  follows  that  taxation,  in  fact,  made  reparation  transfers  possible.  Only,  it  did  so 
in  a  way  very  different  from  that  which  some  economists  thought  of:  they  thought  that 
taxation  would  reduce  system  expenditure  in  Germany,  thereby  depress  the  price  level  and 
thus  produce  the  requisite  export  surplus.  This  it  did  not  do  precisely  because  of  the 
foreign  credits.  But  it  forced  people  to  borrow  abroad  for  domestic  purposes,  and  so 
produced  the  requisite  foreign  exchange  by  a  different  method,  but  for  the  time  being  not 
less  effectively. 


722  BUSINESS  CYCLES 

about,  it  does  not  follow  that  consumers'  welfare  would  have  been,  even 
in  the  short  run,  substantially  impaired. 

If  into  this  picture  we  insert  the  borrowing,  both  domestic  and  foreign, 
of  the  municipalities  and  the  fact  that  much  of  the  expenditure  even  of 
business  concerns  was  unproductive  in  a  commercial  sense,1  we  have 
before  us  the  groundwork  of  the  theory  of  the  specifically  German  pros- 
perity of  the  later  twenties,  which  is  thus  seen  to  link  up  with  the  socio- 
political pattern  of  the  time  in  more  than  one  instructive  way,  and  also 
of  the  specifically  German  form  of  the  subsequent  economic  breakdown, 
as  well  as  of  much  besides.  Some  traits  will  be  added  later;  many  more 
would  have  to  be  added  if  complete  analysis  were  feasible  here.  No 
suggestion  of  a  one-factor  causation  is  intended.  But  our  facts  and  their 
consequences  certainly  suggest  a  lesson. 

5.  England's  postwar  situation  may  usefully  be  compared  to  her 
situation  in  1815.  As  has  been  stated  in  Chap.  VI,  the  national  debt  was, 
in  relative  importance,  similar  in  both  cases  and  so  was  the  depreciation 
of  the  currency  and  the  burden  of  taxation.  The  cases  differ  however  in 
that  the  Napoleonic  wars  were  times  of  vigorous  industrial  and  commer- 
cial expansion  while  the  World  War  left  England  with  her  industrial 
organism  impaired  and  a  loss  of  foreign  investments  which  it  took  her 

1  The  question  of  "productiveness" — or  the  cognate  question:  What  is  to  be  classed  as 
an  "investment"? — is  a  delicate  one  for  other  reasons  besides  the  irritation  it  invariably 
causes.  It  bears  upon  our  subject  in  two  respects.  First,  when  Mr.  Schacht  during  his 
first  tenure  of  office  at  the  Reichsbank  fought  the  rising  tide  of  foreign  debt,  one  of  the 
arguments  was  that  only  such  loans  should  be  permitted  that  would  issue  in  investments 
"productive  of  foreign  exchange."  Whatever  might  be  urged  against  the  correctness  of 
this  argument,  its  real  meaning,  viz.,  that  it  was  dangerous  to  finance  by  foreign  borrowing 
anything  except  commercially  profitable  enterprise,  was  under  the  circumstances  undoubt- 
edly much  to  the  point.  The  municipalities,  at  which  the  Reichsbank's  attack  was  chiefly 
aimed,  had  no  difficulty  in  replying  that  no  less  than  94  per  cent  of  all  their  foreign  loans 
(as  of  Mar.  31,  1928,  see  O.  Mulert  in  the  volume  of  the  Verein  flir  Sozialpolitik  previously 
quoted,  p.  38)  were  in  fact  applied  to  directly  paying  propositions,  electricity  being  the 
biggest  item.  For  municipalities  very  naturally  approached  the  foreign  investor  with  the 
most  businesslike-looking  part  of  their  spending  programs  and  not  with  plans  for  resplendent 
town  halls,  the  funds  for  which  were,  nevertheless,  set  free  by  borrowing  abroad  for  the 
former.  The  productivity  of,  at  all  events,  a  very  large  part  of  the  total  indebtedness, 
domestic  plus  foreign,  both  of  public  and  semipublic  bodies  and  of  private  concerns,  is  in 
fact  open  to  doubt.  Second,  there  has  been  a  tendency  to  define  "investment"  so  exten- 
sively as  to  make  the  term  practically  useless  for  our  purpose.  It  is  owing  to  this  only  that 
the  impressive  figures  of  real  investment  were  arrived  at,  which  have  been  compiled  by  the 
Institut  ftir  Konjunkturforschung  (Sonderheft  22,  1930),  and  which  at  first  sight  seem  to 
contradict  our  diagnosis  of  impoverishment,  excess  consumption,  and  interference  of  tax- 
ation with  the  "formation  of  capital."  To  begin  with,  investment  within  our  meaning 
of  the  term  means  economical  investment  (minimum  outlay  per  unit  of  net  return) :  a 
railroad  is,  no  doubt,  an  economic  enterprise,  but  not  everything  that  is  spent  on  it  is 
therefore  an  investment,  however  wonderful  (and  some  German  station  buildings  are  not 
less  than  wonderful)  it  may  be.  This  applies  even  to  expenditure  conducive  to  excellence 


1919-1929  723 

over  10  years  to  make  good.1  Conquests  of  political  and  economic 
positions  were  made  in  both  cases,  but  net  gains  were  insignificant  in  the 
recent  case  as  compared  to  the  earlier  one.  Also,  practically  the  whole 
world  lay  before  England's  industry  and  trade  in  1815  and  instances  of 
industries  that  had  irreparably  lost  their  previous  positions,  so  con- 
spicuous in  postwar  times  (coal,  cotton  textiles,  shipbuilding),  were 
then  few,  if  not  altogether  absent,  the  only  danger  zone  being  agri- 
culture— which  then  stood  for  the  "key  industries"  that  now  had  to  be 
"safeguarded." 

In  one  respect,  the  frame  of  mind  in  which  England  encountered 
postwar  problems  in  1918  was  curiously  like  that  in  which  she  tackled 
the  situation  of  1815.  In  both  cases,  return  to  the  gold  standard  at  par 
was  considered,  by  dominant  opinion,  as  a  matter  of  course.  The  report 
of  the  Cunliffe  committee,2  which  was  appointed  as  early  as  January 

of  service.  That  is  also  one  of  the  reasons  why  part  even  of  industrial  investment  must 
be  looked  askance  at  from  this  standpoint.  A  large  part  of  the  "investment"  in  agri- 
culture, especially  that  part  which  merely  covered  deficits,  was  clearly  unproductive. 
Moreover,  public  buildings  do  not  constitute  investments,  any  more  than  Louis  XIV  can 
in  any  useful  sense  be  said  to  have  invested  when  he  built  the  palace  of  Versailles.  But 
neither  does  housing  unless  the  dwellings  are  let  at  rents  which  fully  cover  cost,  the  long- 
term  rate  of  interest  included.  More  than  anywhere  else  it  is  here  necessary  to  ward  off 
misunderstanding  by  stating  that  the  writer,  if  he  thought  his  personal  value-judgments 
worth  presenting,  would  have  to  confess  that  he  thoroughly  approved  of  every  single  one 
of  the,  roughly,  11  billion  marks  that  were  spent  on  it  from  1924  to  1928  (of  which  more  than 
half  was  from  public  sources,  4  billions  being  loans  at  3  per  cent  financed  by  a  housetax 
levied  from  houses  which  inflation  had  freed  from  debt,  the  llauszinssteucr  introduced  by 
the  emergency  decree  of  Feb.  14,  1924).  This,  however,  is  entirely  irrelevant  to  our  dis- 
cussion. Relevant  is  merely  that  this  was  largely  consumers'  expenditure  producing  the 
effects  of  consumers'  expenditure.  Thus  corrected,  the  impressive  total  of  net  investment 
amounting  to  about  39.5  billions,  which  the  Berlin  Institute  presents  for  1924  to  1928, 
dwindles  to  considerably  less  than  half.  It  should  be  added  that  increase  in  inventories 
should  be  included  only  insofar  as  it  is  not  due  to  inability  to  sell  and  that  the  figures  for 
internal  accumulation  of  concerns  (undivided  profits)  must  be  interpreted  in  the  light  of 
the  fact  that  there  was  a  systematic  downward  bias  in  depreciation  accounts,  because, 
after  inflation,  the  new  gold  values  were  very  generally  put  at  figures  which  were  too  low. 
The  statement  to  be  found  in  a  recent  study  that  two-thirds  "of  the  [foreign]  capital 
supplies  made  available  to  the  German  economy  in  1924  to  1928  •  •  •  were  applied  to 
extension*of  plant  and  equipment"  is  hence  seen  to  be  misleading.  A  good  discussion  of 
facts  and  problems,  not  quite  however  in  accord  with  the  views  expressed  in  this  note,  as 
well  as  most  of  the  relevant  literature  and  material,  will  be  found  in  E.  Welter,  Die  Ursachen 
des  Kapitalmangels  in  Deutschland,  1931. 

1  Great  Britain's  foreign  investments  reached  the  prewar  figure,  roughly  4  billion 
pounds,  in  1931,  but  even  then  the  presence  of  about  240  millions  of  short  balances  must 
be  set  against  it.     Also  a  significant  change  had  occurred  in  the  character,  particularly  the 
productive  possibilities,  of  those  investments. 

2  The  laconic  Bradbury  report  (Committee  on  the  Currency  and  Bank  of   England 
Note  Issue,  1925)  brushed  aside  all  questions  of  principle  and  can  only  be  described  as  a 
practical  clincher. 


724  BUSINESS  CYCLES 

1918,  does  indeed  contain  matter  that  would  have  been  uncongenial  to 
some  of  the  authors  of  the  bullion  report,  but  there  is  hardly  a  difference 
in  fundamental  principle  between  the  two  documents.  In  recommending 
speedy  reestablishment  of  the  gold  standard  at  prewar  parity — though 
without  free  coinage  for  private  account — to  be  achieved  by  the  accumula- 
tion of  a  gold  reserve  of  150  million  pounds  and  the  gradual  reduction  of 
the  amount  of  paper  money  in  circulation,  it  expressed  a  view  and  an 
intention  which  at  that  time  undoubtedly  prevailed.  This  %  must  be 
accepted  as  a  datum  of  the  situation.1  In  order  to  understand  it  and  the 
consequences  it  entailed,  two  things  must  be  kept  in  mind.  First,  that 
decision  was  fundamentally  extrarational,  and  all  arguments  that  have 
been  adduced  for  it,  either  from  virtues  of  the  gold  standard  in  general — 
largely  imaginary  under  the  circumstances — or  from  the  particular 
interests  of  England  as  the  world's  banker — which  position  was  partly 
untenable  in  any  case  and,  moreover,  as  much  endangered  as  buttressed  by 
that  policy — were  no  more  than  ex  post  rationalizations  of  what  really 
was  a  foregone  conclusion  and  to  many  minds  involved  the  national 
honor  or,  at  least,  prestige.  Second,  the  public  and  especially  those 
labor  men  who  favored  that  policy  were  certainly  not  aware  of  the 
sacrifices — great  or  small — which  in  the  short  run  it  was  bound  to  entail. 
And  those  responsible  men  who  may  reasonably  be  assumed  to  have  been 
aware  of  them  seem  to  have  overlooked  that  1918  was  not  1816,  i.e., 
the  fact  that  their  policy  would  have  to  work  in  an  uncongenial  social 
environment.  What  the  public  wanted  and  what  the  "  responsibles  " 
were  driven  back  upon  thus  amounted  to  an  attempt  to  swim  the  channel 
without  getting  wet.  Perhaps  it  is  surprising  that  the  impossibility  of 
playing  the  orthodox  game  just  in  one  sector  of  national  policy  while 
it  was  clearly  up  in  every  other  should  not  have  occurred  to  bankers  and 

1  By  considering  that  volonte  generate  as  a  datum,  the  writer  merely  wishes  to  stress  a 
fact  which  is  relevant  to  his  narrative.  But,  as  the  text  will  amply  show,  he  has  no  wish  to 
"justify"  it.  Since  many  eminent  English  economists  feel  so  strongly  on  this  and  cognate 
points,  it  will  be  conducive  to  better  understanding  to  state  expressly  that  the  writer 
largely  agrees  with  the  advice  and  criticism  offered  by  some  of  them  at  the  time  and  later, 
in  particular  with  the  practical  upshot  of  Mr.  Hawtrey's  arguments.  As  will  also  be  seen, 
he  rates  at  a  lower  value  both  the  effects  of  the  monetary  policy  pursued  and  the  possi- 
bilities of  available  alternatives.  And  he  does  not  think  it  irrelevant  to  speak  of  the 
effects  that  policy  might  have  had  if  other  elements  of  the  economic  and  social  pattern 
had  shaped  differently.  But  within  a  clausula  rebus  sic  stantibus  he  believes  Mr.  Hawtrey 
to  have  been,  in  the  main,  right  as  far  as  practical  policy  is  concerned.  The  opportunity 
may  be  taken  to  add  an  analogous  statement  with  respect  to  other  matters  of  England's 
economic  policy,  which  have  to  be  touched  in  passing.  In  particular,  there  is  not  only  no 
intention  to  attack  the  policies  recommended  by  a  brilliant  group  of  English  fellow  econo- 
mists, but  iri1  many  if  not  in  all  cases  the  writer  entertains  no  doubt  whatever  about  the 
wisdom,  ex  visit  of  English  short-run  interests,  of  the  advice  proffered  by  them.  This  is 
perfectly  compatible  with  great  differences  in  theory  and  diagnosis. 


1919-1929  725 

politicians.  Perhaps  it  is  not.  But  in  any  case  the  ultimate  failure 
must  be  understood  in  this  light  and  not  in  the  light  of  any  general  merits 
or  demerits  of  such  a  policy,  which  are  an  altogether  different  matter. 

Actual  "deflation"  (in  the  sense  of  reduction  of  the  amount  of  cir- 
culating medium  without  a  corresponding  reduction  in  physical  volume 
of  transactions)  was  attempted  but,  meeting  with  resolute  resistance  to 
the  necessary  adjustments,  was  given  up  quickly.  All  the  government 
did  beyond  liquidating  war  expenditure  was  to  reduce  the  circulation  of 
currency  notes.1  The  Bank  continued  its  efforts  to  accumulate  gold. 
Its  stock  of  coin  and  bullion  fell  slightly  in  1922  but  almost  reached  the 
Cunliffe  goal  by  1925  (at  the  end  of  the  year,  144.6  millions)  and  surpassed 
it  in  1926  (at  the  end  of  the  year,  151.12  millions).  Notes  in  circulation 
roughly  moved  in  step  until  1927.  Other  Securities  were  close  to  their 
1919  figure  both  in  1924  and  1925.  Other  Deposits  fluctuated  strongly 
around  a  level  that  declined  but  little  from  1919  to  the  end  of  1921. 
Total  Clearings  fell  in  1921,  when  Deposits  of  London  Clearing  Banks  were 
still  increasing,  and  began  to  recover  toward  the  end  of  1923  (see  below, 
Sec.  F).  This,  for  the  moment,  is  all  the  background  we  want.  It  also 
suffices  to  assist  the  reader  in  forming  an  opinion  how  much  the  monetary 
element  as  such  can  possibly  have  had  to  do  with  the  crisis  of  1921  and 
the  slowness  of  the  recovery  in  the  two  subsequent  years.  So  far  the 
analogy  with  the  course  of  events  from  1815  to  1821  stands  out  suggestively. 

Meanwhile,  the  pound  was  practically  left  to  follow  its  course.  On 
the  cessation  of  American  pegging,  it  had  begun  to  fall,  discount  reaching 
its  high  point,  34  J^  per  cent,  in  February  1920.2  But  it  speedily  recov- 
ered and  hovered  around  a  10  per  cent  discount  in  the  summer  of  1924. 
This  recovery  cannot  to  any  great  extent  have  been  due  to  the  actual 
measures  (excluding  speeches)  taken  by  government  or  the  Bank.  Nor 
can  it  be  fully  accounted  for  by  the  fall  in  price  level  that  had  occurred 
but  in  the  main  was  an  international  phenomenon.  Much  more  impor- 
tant was  that  same  factor,  which  subsequently  also  made  the  last  step 
so  temptingly  easy:  by  that  time  the  whole  world  was  expecting  that 
England  would  return  to  the  prewar  parity  and,  hence,  was  .buying 
sterling  exchange.  The  temptation  proved  irresistible:  the  hope  that 
gold  would  internationally  fall  to  meet  the  pound  was,  to  say  the  least, 
vague;  the  domestic  situation  offered  but  little  ground  for  optimism;  real 
success  in  the  sense  of  achieving  a  gold  standard  that  would  normally  work 

1  There  were  367.6  million  pounds  of  them  in  1920  and  295.6  millions  in  1925.     Later 
measures  dealing  with  them  need  not  be  discussed  in  this  book. 

2  An  official  embargo  was  laid  on  gold  and  silver,  but  it  merely  continued  a  state  of 
things  which,  by  patriotic  discipline,  had  obtained  throughout  the  war.     After  that  official 
embargo  had  been  lifted,  attempts  were  made  amidst  the  difficulties  that  ensued  to  resort 
again  to  the  latter  method  by  frowning  on  gold  exports,  mobilizing  public  opinion  against 
operators,  inducing  shipping  companies  to  raise  freights  for  gold,  and  so  on. 


726  BUSINESS  CYCLES 

at  prewar  parity  was  for  the  moment  out  of  the  question;  but  technical 
success  was  within  easy  reach.  Short  balances  flocked  to  England 
to  profit  from  a  rise  that  was  a  practical  certainty,  and  by  December  1924 
the  pound  was  no  more  than  1J^  per  cent  below  parity.  The  only  diffi- 
culty was  in  the  dangerous  pressure  which  was  to  be  expected  the  very 
moment  parity  would  be  reached.  For  then  speculation  would  naturally 
realize  and  withdraw.  Against  this  danger  several  defenses  were  built  up. 
An  American  credit  constituted  one  safeguard  to  be  used  if  necessary 
after  the  event.  The  second  safeguard  was  a  bank  rate  of  5  per  cent, 
which  the  Bank  was  careful  to  make  effective.1  The  third  consisted  in 
keeping  the  pound  down  until  as  nearly  as  possible  the  moment  of  the 
plunge.  This  was  successfully  done  by  buying  up  foreign  exchange, 
ostensibly  for  the  purpose  of  servicing  the  debt  to  the  United  States. 

Once  we  accept  the  goal,  nothing  but  admiration  can  be  felt  for  what 
once  more  was  a  very  fine  piece  of  steering,  that  not  only  achieved 
technical  success  but,  in  doing  so,  also  avoided  jerks  and  jolts  and  mini- 
mized injury  to  the  economic  organism.  Reasons  will  be  offered  (Sec.  F) 
for  believing  that  the  Bank's  policy  continued  to  deserve  the  latter  com- 
pliment during  the  subsequent  years.  By  adroit  use  of  gold  devices  and 
masterly  handling  of  the  short-loan  market  it  undoubtedly  made  the  best 
of  very  delicate  situations.  It  is  nonetheless  clear  that  the  Gold  Standard 
Act  marks  not  the  end  but  the  beginning  of  the  real  difficulties.  On  the 
face  of  it,  the  new  gold  standard  was  untenable  and  bound  to  break 
down — if  worked  according  to  classical  principles — through  an  unavoid- 
able efflux  of  gold.  Obviously,  the  Bank  and  the  government  must  have 
hoped  that  it  would  be  possible  for  a  time  to  maneuver  so  as  to  avoid  the 
latter  and  that  during  that  time  domestic  or  international  developments 
would  resolve  the  dilemma.  The  first  proved,  in  fact,  possible,  the  latter 
•failed  to  mature.  It  is  at  this  point  that  the  difference  begins  between 
the  course  of  events  after  the  Napoleonic  war  and  the  course  of  events 
after  the  World  War  of  our  time.  Then,  drastic  readjustment,  by  means 
of  monetary  policy,  of  the  price  level  and  of  incomes  would  have  been 
possible,  but  vigorous  evolution  of  the  national  economy  made  it  unneces- 
sary. Now,  no  comparable  development  setting  in,  drastic  readjustment 
by  monetary  policy  would  have  been  necessary,  but  it  was  not  possible. 
The  very  fact  that  substantially  similar  monetary  policies  produced  such 
different  results  and  that  hardly  any  difficulties  of  a  purely  monetary 
kind  were  encountered  in  the  one  case,  while  difficulties  of  this  nature 
eventually  proved  insuperable  in  the  other,  should  convince  anyone  that 
monetary  policy  has  no  claim  in  a  general  theory  to  the  key  position 
1  Great  funding  operations  coupled  with  the  reissue  of  about  the  same  amount  of 
treasury  bills  achieved  that  end.  On  the  other  hand,  banks  were  made  to  understand  that 
foreign  lending  was  not  for  the  time  being  considered  as  in  the  national  interest  ("embargo 
on  capital  export"). 


1919-1929  727 

allocated  to  it  by  some  economists.  It  is  no  contradiction — but,  on  the 
contrary,  a  trivial  corollary — to  say  that  if  all  the  other  elements  of  the 
social  and  economic  setup  are  taken  as  given  and  if  we  are  left  with 
monetary  policy  as  the  only  variable,  the  latter  will  acquire  a  causative 
importance  and,  in  particular,  become,  if  incompatible  with  the  rest  of 
the  setup,  a  depressing  external  factor.  This  is  what  happened  in  our 
case  and  what  defines  the  role  of  England's  monetary  policy  in  her  post- 
war cyclical  process.  Overvaluation  of  the  pound — putting  a  bounty  on 
imports  and  penalizing  exports — and  a  bank  rate  abnormally  high  under 
the  circumstances  are  two  familiar  instances  of  its  modus  operandi, 
although  we  should  not  overestimate  the  importance  of  either. 

Thus  our  argument  leads,  in  this  sense,  to  agreement  with  those 
English  authorities  on  money  who  hold  that,  all  other  things  being  as 
they  were,  the  return  to  the  gold  standard  or  the  return  to  it  at  prewar 
parity  spelled  pressure  that  aggravated  difficulties  and  could  have  been 
alleviated  by  another  monetary  policy  which  in  turn  need  not  have 
produced  other  difficulties.  Much  more  important  for  explanation, 
however,  although  under  the  circumstances  perhaps  not  for  practical 
advice,  were  some  of  those  other  things.  Neither  monetary  policy  nor 
the  fact  that  England's  domestic  and  international  position  was  not  so 
favorable  in  1918  as  it  was  in  1815  will  sufficiently  account  for  what  is 
universally  felt  to  be  unsatisfactory  economic  performance.  The  funda- 
mental social  change,  which  must  destroy  the  frame  and  atmosphere 
conducive  to  the  working  of  the  capitalist  engine  at  maximum  efficiency 
and  which  we  have  tried  to  analyze  in  Sec.  B,  is,  with  the  exception  of 
Russia,  nowhere  so  obvious  as  it  is  in  England.  The  difference  between 
what  Parliament  did  with  the  income  tax  respectively  in  1816  and  in  1918 
is  a  significant  symptom  even  for  those  who  refuse  to  look  upon  it  as  a 
cause.  England's  fiscal  policy1  characterizes  a  social  situation  which 
hardly  displays  any  symptoms  that  might  be  interpreted  in  the  opposite 
sense. 

What  is  typically  English  in  this,  however,  and  what  strikes  any 
outside  observer  more  than  anything  else  is  the  unrevolutionary  form 
of  a  change  which,  among  other  things,  involved  as  great  a  transfer  of 
wealth  as  was  ever  effected  by  any  revolution,  the  Russian  one  alone 
excepted.  By  this  the  writer  does  not  intend  to  refer  only  to  the  absence 
of  a  violent  break  of  legal  continuity  but  to  the  much  more  relevant  fact 

1  As  we  have  seen  in  Chap.  VII,  the  roots  of  that  policy  reach  far  back.  Let  us  recall, 
however,  that  the  earliest  conspicuous  landmark  in  a  long  development  was  the  first 
budget  of  the  Campbell-Bannerman  administration  (Asquith;  Mr.  Lloyd  George's  "people's 
budget"  was  the  second  major  step).  The  essential  point  was  the  earmarking  of  part  of 
the  surplus  that  resulted  from  Boer  War  taxation  for  the  purpose  of  financing  old-age 
pensions  instead  of  applying  it  to  the  reduction  of  taxation  or  debt,  which  would  have  been 
the  "classical"  measure  to  take. 


728  BUSINESS  CYCLES 

of  continuity  in  the  personnel  that  mans  the  political  ship.  The  socio- 
logical pattern  of  that  personnel  changed,  no  doubt,  significantly.  But 
it  did  so  slowly.  And  what  at  any  time  was  the  old  stratum  succeeded  in 
absorbing  new  elements  both  in  the  sense  that  it  readily  received  and  in 
the  sense  that  it  effectively  assimilated  rising  talent.  This  was  possible 
only  because  the  old  stratum  itself — or  its  more  active  elements — had 
an  altogether  unique  ability  to  accept  and  to  handle  fundamentally  new 
situations  and  principles.  The  set  of  people  for  whom  ruling  is  more 
important  than  the  purposes  and  interests  to  be  served  by  ruling  is  in 
England  larger  and  more  influential  than  anywhere  else.  That  set 
carried  free  trade  in  corn,  although  it  was  at  the  time  controlled  by 
agrarian  interests,  and  through  generations  managed  both  domestic  and 
foreign  policy  entirely  from  the  standpoint  of  a  bourgeoisie,  with  which 
it  was  more  or  less  allied  but  by  no  means  identical.  There  is,  thus, 
nothing — short  of  very  great  inconvenience — to  stop  it  from  repeating 
the  feat  and  running  a  Labor  party  or  even,  without  doing  this,  running 
the  country  by  means  of  a  Conservative  party  on  Labor  lines,  scaling  off, 
no  doubt,  rough  edges  but  not  altering  the  fundamental  contour  of  events. 
To  a  certain  extent  this  is  what  actually  happened  in  the  postwar  period,1 
and  the  fact  that  England  emerged  from  the  war  in  a  substantially 
conservative  mood — which  was  intensified,  if  anything,  by  the  social 
unrest,  the  outbreaks  in  England,  and  the  revolution  in  Ireland  in  1920 — 
and  subsequently  kept  Conservative  administrations  in  power  for  most  of 
the  time,  can  hence  not  affect  our  diagnosis  or  its  implications  for  the 
mechanism  or  the  results  of  the  cyclical  process.  Again,  nothing  the 
Labor  party  actually  did,  either  in  office  or  out  of  it,  can  with  any  con- 
fidence be  pointed  to  as  a  major  cause  of  disturbance  deflecting  the  eco- 
nomic process  from  the  course  it  would  otherwise  have  taken.  Its  short 
spells  of  power  and  its  strong  position  in  the  country  are  relevant  for  us 
merely  as  indications  of  the  changes  that  have  occurred  in  social  structure 
and  atmosphere.  In  financial  matters,  in  particular,  it  was  eminently 
"sound'*  in  the  orthodox  sense. 

1  Many  Englishmen  will  not  agree  to,  and  even  be  irritated  by,  the  above  statements. 
But  this  disagreement  is  easily  accounted  for  by  the  difference  in  standpoint  between  the 
man  in  the  thick  of  the  game  and  the  mere  observer.  For  the  one,  the  individual  measure 
matters  in  all  its  details,  and  in  6ghting  for  his  points  he  will  use,  and  have  to  use,  a  stereo- 
typed phraseology  that  knows  no  colors  except  black  and  white,  not  less  but  more  so  when 
the  actual  colors  shade  off  into  each  other.  For  the  other,  nothing  matters  except  broad 
results  and  very  simplified  contours.  Also,  Englishmen  who  struggle  with  the  less  mobile 
elements  of  their  own  environment  will,  no  doubt,  be  unable  to  see  what  strikes  the  writer 
as  extreme  adaptability  to  new  situations  and  principles,  if  principles  be  the  right  word  for 
attitudes  or  opinions  that  are  so  easily  jettisoned  without  sterile  regrets.  Yet  that  adapta- 
bility is  a  provable  as  well  as  most  important  fact.  Peel  and  Disraeli  were  masters  of  an 
art  that  has,  obviously,  not  been  lost, 


1919-1929  729 

War  financing,  too,  had  been  on  classical  lines.  The  writer  finds  it 
difficult  to  understand  the  critics  who  were  not  satisfied  with  what  to  him 
seems  admirable  performance.  The  postwar  budget  of  the  coalition 
government  has,  however,  repeatedly  been  criticized  on  the  score  of — 
epithet  to  be  inserted  by  the  reader,  e.g.t  meritorious,  nefarious,  bene- 
ficial, abominable — extravagance.  This  is  relevant  to  our  subject 
because  it  shows  that  rapidity  of  liquidation  of  war  expenditure  can 
hardly  be  invoked  in  explanation  of  the  subsequent  slump,  except  in  the 
sense  that  continuance  of  war  expenditure  at  an  increasing  rate  would 
have  avoided  it  for  a  time.  Subsequently,  expenditure  was  normalized, 
but  on  a  level  about  four  times  as  high  as  that  of  1913-1914.  The  excess 
profits  tax  was  abolished.  Of  other  adjustments  we  need  mention  only 
the  reduction  in  the  flat  rate  of  the  income  tax  from  72  pence  in  the 
pound,  in  1918-1919,  to  48  in  1925-1926.  Glancing  at  the  state  of  things 
in  the  latter  year  we  find  that  expenditure  of  the  government  as  per 
Finance  Accounts  (which  include  transfers  to  local  authorities  and  pay- 
ments to  North  Ireland,  but  not  other  local  expenditure)  was  826.1  mil- 
lion pounds  and  revenue  812.1,  deficit  turning  into  surplus  if  we  deduct 
the  50  millions  applied  to  the  reduction  of  debt.  This  compares  to 
about  198  million  pounds,  at  which  accounts  roughly  balanced  for  1913- 
1914.  Revenue  of  local  authorities  from  rates  was  166.1  million  pounds 
in  the  later  fiscal  year,  as  compared  with  79  in  the  earlier.  Opinions 
differ  widely  as  to  the  choice  of  an  index  by  which  to  reduce  those  figures 
to  comparability.  It  is  sufficient,  however,  to  state  that  while  total 
money  income  of  the  inhabitants  of  Great  Britain  and  North  Ireland 
about  doubled  between  the  two  years,  total  public  revenue  increased  to 
about  3.4  times  its  former  amount,  which,  looked  at  from  the  standpoint 
of  intact  capitalism,  was  already  extremely  high.  Unlike  German 
public  spending,  that  of  England,  especially  in  its  permanent  elements, 
displayed  some  stability  for  several  years  afterward  and  never  outran 
revenue  to  any  serious  extent.  The  Consolidated  Revenue  and  Expendi- 
ture Account  of  government,  local  authorities,  and  social  insurance 
always  showed  a  surplus.1 

Fiscal  policy  undoubtedly  interfered  with  the  saving-investment 
process,  however.2  This  effect  becomes  still  more  obvious  if  we  take 

1  Cf.  C.  Clark,  National  Income  and  Outlay,  Table  59,  pp.  140  and  141. 

2  According  to  Professor  Bowley  and  Sir  Josiah  Stamp,  The  National  Income  1924 
(1927,  p.  57),  out  of  an  Aggregate  Income  of  2,020  million  pounds  in  1911,  320  millions  were 
saved  and  225  paid  in  rates  and  taxes;  while  out  of  the  1924  aggregate  of  4,165  millions 
475  were  saved  and  855  paid  in  rates  and  taxes.     Very  roughly,  the  amount  "spent  freely" 
increased  as  Aggregate  Income,  i.e.,  doubled;  the  amount  saved  increased  by  one-half,  i.e., 
less  than  price  level;  the  amount  paid  in  rates  and  taxes  increased  to  more  than  8%  times 
its  former  figure.     We  may  differ,  of  course,  as  to  the  effects  on  the  economic  process  of 
this  absolute  fall  in  real  saving  and  this  relative  fall  in  monetary  saving.     But  it  would  be 


730  BUSINESS  CYCLES 

account  of  the  fact  that  the  burden  of  taxation  was  not  only  increased 
but  also  shifted  in  a  way  that  cannot  have  failed  to  affect  saving  and 
accumulation.  Neglecting  the  excess-profits  duty  and  the  tax  on  cor- 
porate business,  taxation  of  inheritance  and  income  yielded  about  74.5 
millions  in  1913-1914  and  about  389  in  1925-1926,  and  though  con- 
sumers' expenditure  of  the  payers  of  these  taxes  was  certainly  curtailed — 
nobody  at  all  familiar  with  English  life  can  have  any  doubt  about  that — 
the  greater  part,  perhaps  two-thirds  of  the  difference  of  £.40  million 
pounds,  which  we  arrive  at  if  we  double  the  prewar  figure  in  order  to 
make  it  roughly  comparable,  must  have  come  out  of  potential  savings  or, 
in  some  cases,  out  of  dissavings.  The  implications  of  this — see  above, 
sub  2  and  passim — may  be  distasteful.  All  that  can  be  replied  is  that 
it  is  really  a  pity  that  facts  have  a  way  of  verifying  views  which  are  so 
obviously  antiquated.  For  the  symptoms  we  simultaneously  observe  in 
the  economic  organism  are  exactly  what  old-fashioned  economists  (whose 
theories  the  writer,  as  this  book  amply  proves,  is  in  general  very  far  from 
sharing)  would  have  expected  to  follow  from  such  a  fiscal  policy. 

Those  implications  are  not  weakened,  first,  by  the  fact  that  relatively 
moderate  but  not  negligible  amounts  were  spent  on  subsidies  to  business 
— agriculture,  air  transport,  the  merchant  marine,  coal  mining,  the  beet 
sugar  and  other  industries.  Nor,  in  the  second  place,  is  it  relevant  that 
other  lines  of  expenditure  will  command  the  support  of  most  of  us  and 
that  still  others  certainly  "increased  productivity."  Policies  of  social 
betterment,  for  example,  accounted  for  about  16  million  pounds  in  1912— 
1913  and  for  about  72  in  1925-1926.  *  Also,  it  must  not  be  forgotten 
that  improvements  of  the  environment,  however  beneficial,  do  in  most 
cases  fail  to  link  up  with  the  tax  that  finances  them  in  such  a  way  as  to 
neutralize  its  economic  effects.  There  is,  for  instance,  no  doubt  that  if 
government  action  involving  expenditure  could  reduce  precipitation  in 
the  Lake  District,  the  nation's  enjoyment  of  that  delightful  country 
would  be  greatly  increased.  But  a  general  tax  levied  for  the  purpose 
would  act  as  a  net  burden  all  the  same,  and  the  benefit  would  simply 
enter  into  the  general  environmental  conditions  exactly  as  if,  instead  of 
H.  M.  Servants,  Jupiter  Pluvius  had  wrought  the  change  gratis. 

highly  unreasonable  to  deny  that  this  fall  itself  was  mainly  due  to  the  increase  in  taxation. 
The  use  of  other  estimates  could  not  affect  the  result  substantially.  We  shall,  however, 
see  reason  to  suspect  (see  infra,  Sec.  F,  II,  1)  that  all  estimates  greatly  exaggerate  the 
amount  of  saving,  which  actually  was  far  smaller  than  475  millions. 

1  Those  are  estimates  of  the  Statistische  Reichsamt,  Staatsausgaben  von  Grossbritan- 
nien,  Frankreich,  Belgien  und  Italien  1927.  A  subjective  element  enters  into  the  various 
definitions  between  which  one  must  choose.  Besides,  the  different  economic  situations 
entailing — also  for  reasons  that  have  nothing  to  do  with  taxation — more  unemployment  in 
the  one  case  than  in  the  other  must  be  taken  into  account.  For  our  argument  all  this 
matters  but  little.  We  are  not  criticizing. 


1919-1929  731 

Finally,  third,  it  must,  in  view  of  a  not  very  creditable  discussion  that 
has  arisen  about  the  point,  be  emphasized  that  taxation  for  the  service 
of  the  5.9  billions  by  which  the  domestic  national  debt  had  been  increased 
in  consequence  of  the  war,  cannot  be  left  out  of  account  on  the  ground 
that  it  effects  a  "mere  transfer."  It  is  true  that  an  internal  debt  in 
some  respects  raises  problems  which  differ  from  those  presented  by  an 
external  one.  It  is  also  true  that  this  type  of  expenditure  does  not  reduce 
the  amount  of  factors  of  production  available  for  industry,  as  expenditure 
on  additional  policemen  would  in  a  case  of  full  employment.  But 
exactly  in  the  same  sense — though  in  no  other — in  which  expenditure  on 
disabled  soldiers  is  a  burden,  expenditure  on  disabled  capital  is,  too. 
Its  effects  on  those  who  pay  the  corresponding  taxes  are  the  same  as  the 
effects  of  taxes  levied  for  any  other  purpose.  And  no  gains  of  either 
class  of  recipients  balance  this  loss:  both  the  disabled  soldiers  and  the 
owners  of  the  "capital'*  which  financed  the  war  live,  without  con- 
tributing, on  the  results  of  the  productive  efforts  of  the  rest  of  the  com- 
munity, while  otherwise  they  would  have  been  earning  wages  and  interest 
by  increasing  the  national  dividend.  No  talk  about  "putting  money 
from  one  pocket  into  the  other"  avails  against  this  fact. 

Other  effects  apart,  this  fiscal  policy  enforced  recourse  to  foreign 
short  balances  in  a  way  that  presents  some  analogy  with  the  German  case. 
The  sober  management  of  the  budget  prevented,  indeed,  a  "consumers' 
prosperity"  in  the  German  sense  from  developing;  and  part  of  the 
English  short  borrowing  links  up  with  long-term  lending  abroad.  But 
it  still  remains  true  that  if  taxation  had  not  cut  so  deeply  into  potential 
savings,  less  foreign  balances  would  have  flowed  in  and  many  things, 
money  rates  among  them,  would  have  been  different  from  what  they  were. 
However,  the  English  situation,  as  pointed  out  above,  more  than  that 
of  any  other  country  was  influenced  by  circumstances  affecting  particular 
spots.  These  circumstances,  such  as  the  innovations  which  depressed 
coal  mining,  or  the  rise  of  native  capitalism  in  India,  which  depressed 
the  cotton  textile  industry,  or  the  losses  England's  international  position 
suffered  in  the  fields  of  banking,  shipping,  and  insurance,  permanently 
altered  the  conditions  of  her  economic  life  and  were  for  her  largely  external 
factors.  But  they  are  much  more  important  in  the  interpretation  of 
her  cyclical  phenomena  than  whether  banks  buy  more  or  less  assets. 
Moreover,  both  her  domestic  and  her  international  environment  were 
bound  to  change  at  an  increasing  rate  in  the  neomercantilist  age.  This 
in  itself  unavoidably  imposed  a  radical  change  of  policy.  Cobden  himself 
could  not  be  accused  of  inconsistency — although,  of  course,  he  might  still 
be  wrong — if  he  now  rose  from  the  dead  to  preach  the  doctrine  that  pro- 
tection, state  enterprise,  and  managed  money  must  be  resorted  to  in  order 
to  transform  the  Empire  into  an  ironclad  world  of  its  own. 


732  BUSINESS  CYCLES 

D.  Outlines  of  Economic  History  from  1919  to  1929. — A  very  rough 
sketch  will  be  sufficient  to  convince  the  reader  that  all  the  major  features 
of  economic  life  during  that  period  in  fact  conform  closely  to  our  idea  of 
a  Kondratieff  downgrade  and  that  none  of  them  fights  against  the  hypo- 
thesis which  this  turn  of  phrase  implies. 

1.  We  begin  with  the  agrarian  sphere.  Both  preceding  Kondratieff s 
displayed  within  their  negative  phases  prolonged  agrarian  depressions. 
We  have  seen  that  in  causation  and  symptoms  they  differed  sufficiently, 
as  between  each  other  and  in  each  case  as  between  countries,  to  cast 
doubt  on  any  very  broad  generalization  about  them,  particularly  with 
respect  to  the  "necessity"  or  "normality"  of  their  occurrence.  We  have 
also  seen,  however,  that  certain  properties  of  Kondratieff  downgrades 
tend  to  produce  depressive  conditions  in  the  agrarian  world  as  a  whole, 
and  that  agricultural  innovations,  if  any,  tend  to  produce  in  sectors  of 
that  world  depressive  conditions  that  may  be  important  enough  to 
create  a  picture  of  general  agrarian  depression.  Obviously,  this  is  what 
we  find  in  the  postwar  period  and  what  provides  the  first  approximation 
into  which  it  is  easy  to  fit  all  the  other  factors  of  agrarian  situations. 
But  the  latter  are,  nevertheless,  important  and  should  not  be  neglected 
merely  for  the  sake  of  one-factor  theories  and  one-remedy  therapeutics. 

Primarily,  the  fall  in  agrarian  prices  was  a  fall  not  in  relative  but  in 
absolute  price,  i.e.,  an  element  of  the  fall  in  the  general  price  level.  Such 
a  fall  is  part  of  the  mechanism  of  cyclical  downgrades — of  Kondratieff 
downgrades,  in  particular.  As  we  have  seen,  it  would  not,  in  itself, 
suffice  to  produce  an  agricultural  crisis,  although  it  may  adversely  affect 
the  welfare  of  the  agrarian  community  if  the  farm  prices  of  products 
fall  more  than  the  retail  prices  of  the  finished  products  which  it  buys.1 
"Crisis'*  may  ensue,  however,  if  the  fall  of  the  price  level  impinges  on  a 
debt  situation  that  has  developed  from  borrowing  either  for  unproductive 
purposes — such  as  the  acquisition  of  land — or  for  insufficiently  productive 
ones — such  as  mere  expansion.  But  in  the  United  States  and  England 
agriculture  had  to  face,  as  it  had  after  the  Napoleonic  wars  and,  in  this 

1  This,  of  course,  is  invariably  the  case.  As  to  farm  machinery,  there  is  an  investigation 
by  the  Federal  Trade  Commission  (1920),  which  found  that  prices  of  farm  implements 
were  advanced  by  producers  and  dealers  in  1917  and  1918  by  more  than  was  "warranted  by 
the  increase  in  their  costs,"  especially  in  those  lines  which  were  practically  controlled  by  one 
concern — mowers  and  binders,  for  instance — and  on  which  the  "premium  for  innovation" 
was  more  completely  collected  than  on  others.  But  it  is  true  in  general,  that,  partly  owing 
to  local  dealers*  margins  and  to  freight,  which  together  amount  in  some  cases  to  about  25 
per  cent  of  the  price  paid  by  the  farmer,  partly  to  monopolistically  competitive  situations, 
prices  of  these  things  tend  to  be  both  high  and  rigid.  Servicing,  which  the  farmer  cannot 
as  a  rule  do  himself,  is  also  expensive.  The  same  applies  to  most  of  the  gadgets  of  modern 
life  which  the  farmer  buys  as  a  consumer  and,  to  a  lesser  extent,  to  finished  industrial  products 
generally,  although  some  of  these  are  supplied  to  him  efficiently  and  cheaply  and  without 
any  abnormal  profit. 


1919-1929  733 

country,  the  Civil  War,  not  that  kind  of  fall  in  price  level  which  is  a 
normal  element  of  the  'economic  process  in  Kondratieff  recessions  and 
depressions,  but  the  much  more  violent  reaction  of  prices  to  the  rise  dur- 
ing the  World  War.  Moreover,  agriculture  had  been,  to  the  extent 
indicated  in  the  last  section  of  Chap.  VII,  an  innovating  industry,  or 
rather  an  industry  that  had  innovations  forced  upon  it  which  originated 
elsewhere,  such  as  the  internal-combustion  engine,  specifically  agri- 
cultural machinery,  electric  power  and  appliances,  new  fertilizers.  As 
we  should  expect,  these  innovations  fully  conquered  and  came  to  fruition 
in  the  downgrade,  and  they,  as  well  as  the  locational  shifts,  which  con- 
stitute the  most  important  of  agriculture's  own  innovations,  sectionally 
reduced  costs  to  a  level  on  which  large  sectors  were  unable  to  compete: 
the  food  problem  of  humanity  was,  as  far  as  the  economic  process  was 
concerned,  indeed  definitively  solved,  but  at  the  expense  of  the  agricul- 
tural interest.  Competition  by  other  countries,  development  of  which 
was  accelerated  by  the  war,  harvests,  conditions  of  demand,1  international 
barriers,  and  other  factors  have  to  be  inserted,  however,  to  complete  the 
picture  as  it  unfolded  itself  from  year  to  year. 

a.  Elaborating  a  little  for  each  of  our  countries,  we  will  first  notice 
that  in  the  United  States  the  Bureau  of  Agricultural  Economics  index 
of  prices  received  for  farm  products  rose  from  1915  to  1919  by  109  per 
cent,  while  the  index  of  commodities  bought  by  farmers  rose,  until  1920, 
by  94  per  cent.2  The  year  1920  brought  a  moderate  fall  and  1921  a 
fall  to  116  per  cent  of  the  prewar  figure,  from  which  the  index  of  farm 
prices  recovered  quickly,  to  reach  a  peak  of  147  per  cent  in  1925.  Then  it 

1  The  above  refers  to  the  facts  that  demand  tended  to  become  inelastic  in  some  countries, 
particularly  in  the  United  States,  and  that  the  impulse  that  had  caused  it  to  shift  favorably 
all  through  the  prewar  time  slackened  in  others  (not  in  the  United  States).     Also,  new 
modes  of  life  engendering  new  tastes  and  habits  tended  to  reduce  consumption  of  the  heavier 
foods  per  head.     But  these  facts  have  nothing  to  do  with  any  shortage  of  either  real  or 
monetary  purchasing  power  of  consumers.     This,  as  the  course  of  wages  in  our  three 
countries  amply  proves,  is  again — still  more  obviously  than  it  was  in  the  two  other  instances 
— a  pure  myth,  except  in  the  cases  of  Germany  and  Austria  during  their  inflations.     Pro- 
tectionist policies  have  more  claim  to  our  attention. 

2  These  indices  are  reproduced,  for  example,  by  E.  G.  Nourse  in  Recent  Economic 
Changes,  vol.  II,  1929,  p.  548.     Newer  and  more  extended  investigations  have  not  greatly 
altered  the  general  picture.     It  is  hardly  necessary  to  point  out  the  limits  of  the  reliability 
of  such  indices,  particularly  of  the  nation-wide  type.     Since  interest  on  mortgages,  rail- 
road fares,  and  other  items  on  which  farmers'  receipts  are  spent  did  not  move  at  all,  or 
moved  less  than  the  index  (taxes  on  farm  property  were  in  1919,  according  to  the  same 
source,  only  30  per  cent  above  1914),  the  case  in  1919  was  still  more  favorable  than  it 
seems  to  be  at  first  sight.     This  only  accentuates  the  subsequent  reversal.     Taxes  on  farm 
property  rose  within  our  period  to  over  two  and  a  half  times  the  prewar  figures,  and  the 
index  of  prices  of  the  commodities  bought  by  farmers  never  fell  below  about  150  per  cent 
of  those  figures.     Discontent  was,  therefore,  understandable,  especially  if  we  take  account 
of  the  rapid  increase  in  wealth  in  the  industrial  sector.     But  discontent  is  not  crisis. 


734  BUSINESS  CYCLES 

fluctuated  on  a  moderately  falling  "trend"  up  to  the  eve  of  the  world 
crisis.  This  development  must  be  correlated  with  the  development  of  the 
agricultural  debt. l  Even  in  the  (predominantly  prosperous)  prewar  years 
total  farm  mortgages  were  considerable — 3.3  billions  in  1910.2  They  rose 
to  237  per  cent  of  that  sum  by  1920,  quite  enough  to  produce  many  unten- 
able situations,  even  if  we  take  into  account  the  fact  that  incomes  had 
risen  more  than  farm  prices,  and  afterward  fell  less  than  they  did.  But  to 
1925  there  was  a  further  increase  to  about  9.36  billions,  the  peak  that 
occurred  in  1928  being  but  insignificantly  higher — 9.46.  Now,  some 
part  of  this  load  probably  was  the  result  of  funding  short-time  debts 
which  had  become  irksome,  and  a  greater  part,  the  result  of  expansion 
and  of  the  mechanization  to  be  mentioned  presently.  But  the  correlation 
of  the  two  periods  of  increase  with  rising  land  values  is  obvious,  and  the 
inference  is  unavoidable  that  much  of  this  increase  in  debt  came  from 
purchases  of  land  with  a  view  to  reaping  not  harvests  but  increments  of 
value.3  So  far,  then,  we  conclude:  there  was  a  short  and  sharp  crisis  in 
agriculture  in  1920—1921,  which  was  part  of  the  general  postwar  slump, 
though  accentuated  for  agriculture  through  the  burden  of  partly  unpro- 
ductive debt.  In  the  years  to  1926  there  was,  however  unsatisfactory 
the  situation  may  have  been  from  other  standpoints  than  ours,  no  general 
agrarian  depression  at  all.  After  1926  and  to  the  threshold  of  the  world 
crisis,  the  agrarian  situation  became  increasingly  unsatisfactory,  but  the 
only  general  cause  of  this  was,  again,  the  pressure  of  unproductive  debt. 
But  this  diagnosis  misses,  besides  many  minor  points,  a  major  one, 
viz,,  the  influence  of  the  innovations  mentioned  before.  Some  of  them, 
like  the  progress  in  the  cultivation  of  citrus  fruit  and  vegetables  or  in 
refrigeration  and  canning,  did  not — or  not  materially — spell  competition 
of  some  sectors  or  products  with  others,  and  brought  a  net  addition  to  the 

1  Mortgages  were,  of  course,  only  part  of  the  total  agrarian  debt.     But  since  part  of  the 
short  debts  were — in  cattle  ranching,  for  instance — really  for  the  purposes  of  current  busi- 
ness and  not  dangerous,  we  can  confine  ourselves,  in  a  sketch  of  this  kind,  to  the  mortgages. 

2  Figures  on  indebtedness  are  taken  from  the  publication  of  the  Bureau  of  Foreign  and 
Domestic  Commerce  on  Long-term  Debts,  1937,  p.  107. 

8  Official  publications  are  understandably  careful  in  dealing  with  that  delicate  point. 
Still,  the  publication  quoted  in  the  preceding  note  goes  so  far  as  to  say  (p.  106) :  "The  war 
and  postwar  period  of  rising  land  values  and  enlarged  volume  of  land  transfers  resulted  in 
a  marked  increase  in  the  volume  of  farm  mortgage  debt,  as  mortgages  were  freely  used  to 
facilitate  sales."  In  a  country  in  which  everyone  speculated,  there  should  be  no  shame- 
facedness  about  this.  Remedies  are  not  found  by  concealing  the  patient's  ills.  Moreover, 
although  in  some  states  that  "free  use  of  mortgages"  went  rather  far,  it  is  not  held  that  the 
speculation  in  farm  land  in  the  country  as  a  whole  went  to  anything  like  the  lengths  it  did, 
for  instance,  in  New  Zealand.  In  Mississippi,  83.7  per  cent  of  farms  were  mortgaged;  in 
West  Virginia,  only  32.6  per  cent.  The  point  under  discussion  has  been  officially  recog- 
nized. See,  for  example,  Secretary  H.  A.  Wallace's  statements  in  The  Agricultural  Situa- 
tion, issued  by  the  Bureau  of  Agricultural  Economics,  June  1,  1937. 


1919-1929  735 

total  of  agricultural  incomes.  To  a  lesser  extent  this  is  true  also  of 
poultry  and  cattle  rearing  and  of  dairying. 1  Others,  like  some  electrical 
appliances,  even  helped  sectors  that  were  being  put  out  of  existence  by 
competition,  especially  those  whose  main  difficulty  was  dear  labor.  But 
most  of  the  improvements  in  the  methods  of  agriculture,  while  instru- 
mental in  bringing  forth  agriculture's  share  in  that  rising  tide  of  con- 
sumers' goods  which,  according  to  our  schema,  is  a  feature  of  Kondratieff 
downgrades,  and  even  producing  agricultural  prosperity  in  wide  sectors 
of  the  country,2  tended  to  push  certain  regions  below  the  margin  of 
profitable  production.  This,  of  course,  is  wholly  true  of  the  productive 
success  achieved  on  reclaimed,  drained,  irrigated  lands  and  of  the  process 
by  which  large  areas  have  been  taken  into  intensive  cultivation  of  crops 
and  into  horticulture  that  previously  served  purposes  of  "extensive" 
farming.  But  it  is  partly  true  also  of  the  truck,  the  tractor,  most  of  the 
machinery  newly  introduced  into  grain  farming,  and  to  some  extent  of 
the  use  of  electrical  power.  Most  of  them  increase  the  optimum  size 
of  the  farming  unit,  some  of  them  can  be  used  to  full  advantage  only 
under  the  particular  conditions  of  the  Great  Plains.  From  1920  to  1930 
the  number  of  motor  trucks  increased  from  139,000  to  900,000  and  the 
number  of  tractors  from  246,000  to  920,000.3  The  latter  invites  the 
combination  of  operations  that  were  previously  quite  distinct,  ploughing 
and  the  preparation  of  the  seedbed,  for  instance,  and  thus  steadily  leads  to 
ever-increasing  mechanization.  The  use  of  the  combine  harvester,  which 
had  first  been  a  success  in  California,  spread  and  yearly  sales  increased 
nearly  sevenfold  in  the  same  period.  Cotton-  and  corn-harvesting  imple- 
ments must  be  added,  but  no  further  examples  are  necessary  in  order  to 
establish  our  point.  Nothing  of  all  this  was  fundamentally  new;  all  of 
it  is  typically  "induced  development"  of  the  kind  which  on  previous 

1  An  eminent  politician,  instead,  professed  himself  shocked  at  finding  that  Wisconsin 
milk  and  milk  products  are  being  sold  in  Georgia;  but  neither  farmers  nor  economists  are 
likely  to  share  the  feelings  he  obviously  entertains  on  the  subject  of  national  division  of 
labor. 

2  That  prosperity  was,  however,  less  accentuated  than  might  have  been  expected,  not 
only  because  profits  had  to  be  shared,  as  pointed  out  before,  with  the  industries  that  were 
responsible  for  the  innovations,  but  also  because  of  the  perfectly  competitive  character  of 
agriculture,  which  responded  to  the  lowering  of  costs  by  a  prompt  reduction  of  the  prices 
of  products.     This  is  what,  together  with  the  fact  that  the  undersold  units  did  not  promptly 
disappear,  created  that  impression  which  is  sometimes  conveyed  by  the  phrase  "agri- 
cultural overproduction."     If  the  implication  is  that  farmers  are  to  "blame"  for  their 
plight,  the  reply  is  in  order  that,  by  behaving  as  they  did,  they  fully  rendered  the  social 
service  that  circumstances  enabled  them  to  render.     In  any  case,  however,  that  phrase  is 
misleading,  for  the  state  of  things  envisaged  was  not  simply  overproduction  in  the  ordinary 
sense  of  the  word. 

3  The  low-priced  tractor  (Fordson,  International  Harvester),  in  particular,  did  not  come 
in  before  1915. 


736  BUSINESS  CYCLES 

occasions  we  found  to  be  characteristic  of  Kondratieff  downgrades.  Our 
old  formula,  Depression  spotted  by  Prosperity,  fits  the  case  as  it  did  the 
others.1  Emigration  from  agriculture  to  industry  was,  under  the  circum- 
stances, from  the  standpoint  of  the  logic  of  the  capitalist  mechanism,  a 
perfectly  normal  phenomenon  of  adaptation. 

Other  aspects  of  the  same  features  and  some  additional  features  of 
the  postwar  agrarian  situation  will  come  into  view  if  we  glance  for  a 
moment  at  cotton  and  wheat  in  particular.2  Ever  since  the  beginning 
of  the  nineties  the  price  of  cotton  moved  fairly  well  with  any  all-com- 
modity index,  domestic  consumption  of  cotton — also  roughly — with  any 
index  of  industrial  production.  This  on  the  whole  remained  so  in  our 
period,  the  chief  exception  being  the  rapid  recovery  of  cotton  consumption 
in  1921,  right  from  the  beginning  of  the  year.  Quantity  exported  was 
below  the  average  of  the  last  prewar  years  in  1922  to  1924,  but  roughly 
on  the  same  level,  or  somewhat  above,  in  1924  to  1929,  value  rising 
sharply  from  1921  to  1925  and  receding  afterward.  Rayon  was  only  one 
of  the  competing  commodities  that  must  have  exerted  some  influence — 
with  increasing  wealth,  the  competition  of  wool  increases  in  many  lines — 
but  owing  to  the  emergence  of  new  uses,  such  competition  was  of  but 
minor  importance;  the  standard  fiber  was  still  to  come.  There  was  the 
migration  to  lands  made  available  by  new  methods  of  cultivation, 
especially  to  Texas  and  Oklahoma,  partly  due  to  the  tractor  and  the 
mechanical  picker  (complemented  by  a  corresponding  innovation  in 
ginning),  with  a  consequent  competitive  annihilation  of  much  of  the 
Southeastern  cotton  farming. 

In  all  this  our  process  shows  to  perfection,  and  the  process  of  labor 
being  drawn  away  from  an  "old"  stratum  (toward  the  Northeastern 
industry)  is  particularly  in  evidence.  Farmers'  price  of  the  standard 

xlt  follows,  of  course,  that  "price  parities"  between  agricultural  and  nonagricultural 
products  are  irrelevant  for  an  appraisal  of  the  degree  of  welfare  enjoyed  by  the  agricultural 
community.  But  "income  parities,"  as  between  the  agricultural  community  as  a  whole 
and  the  industrial  sector,  have  not  much  more  economic  meaning.  There  is  not  more 
reason  to  expect  agricultural  prices  or  incomes  to  keep  a  certain  relation  to  other  prices  or 
incomes  than  there  is  to  expect  that,  say,  coal  and  income  derived  from  coal  mining  should 
permanently  keep  its  relative  position  in  an  economic  world  that  is  incessantly  being 
revolutionized.  If  for  extra-economic  reasons  agriculture  is  to  be  subsidized — and  there 
are  many  such  reasons — it  should  be  done  without  reference  to  any  such  parities  between 
the  present  and  a  bygone  state  of  things.  But  total  market  value  of  farm  products,  while 
it  fell  much  more  than  money  income  of  industrial  workers  in  1920,  from  that  year  on  sub- 
stantially kept  pace  with  it.  See  M.  Ezekiel,  Evaluating  1933  for  the  Farmer,  Journal 
of  the  American  Statistical  Association  for  June  1934,  chart  on  p.  140. 

2  The  same  phenomena  would  stand  out  still  better  in  other  instances,  such  as  rubber, 
coffee,  and  rice,  which  are,  however,  not  products  or  not  leading  products  of  any  of  our 
countries.  The  studies  on  the  Artificial  Control  of  Raw  Material  Supply  published  in  the 
London  and  Cambridge  Economic  Service  (J.  W.  F.  Rowe)  should  be  mentioned. 


1919-1929  73T 

quality  rose  from  about  12^  cents  to  about  28^  cents  during  the  war 
(the  latter  figure  being  of  November  1918),  which  was  perfectly  normal 
and  neither  justified  nor  actually  induced  increase  in  acreage.  Acreage 
harvested  actually  fell  from  its  1914-1915  peak.  The  ravages  of  the 
boll  weevil  in  1921,  1922,  and  1923,  however,  raised  it  to  32  cents  toward 
the  end  of  the  latter  year,  and  this  presumably  propelled  the  expansion 
in  the  West,  which  in  spite  of  abandonments — not  all  due  to  the  boll  , 
weevil — had  set  in  before  and  carried  total  acreage  in  this  country  from 
the  29.7  million  acres  in  1921  to  about  45.8  in  1926.  Acreage  outside  the 
United  States  at  the  same  time  increased  from  about  28.5  to  over  42  mil- 
lion acres,  not  only,  of  course,  in  response  to  that  price,  but  in  consequence 
of  the  endeavor  made  in  many  countries  to  develop  cotton  growing, 
which  date  far  back  into  prewar  times  and  were  indirectly  fostered  by 
the  tariff  policy  of  the  United  States.  Thus  there  developed  slowly, 
beneath  the  surface  of  current  fluctuations,  an  untenable  situation  which 
was  bound  to  curtail  the  role  of  American  cotton  in  the  world  and  to 
explode  in  a  major  depression.  The  presence  of  overproduction  in  the 
proper  sense  of  the  term  is,  in  this  special  case,  as  undeniable  as  the 
rationale  of  the  argument  for  planned  retreat.1  Cottonseed  oil  and  its 
residues  cannot  be  dealt  with  here,  but  their  possibilities  in  the  fields  of 
human  and  animal  food  and  of  chemical  products,  although  very  con- 
siderable, would  not  fundamentally  alter  the  picture. 

The  postwar  wheat  situation  presents  fundamentally  the  same  fea- 
tures, yet  differs  in  important  respects.  Before  the  war,  United  States 
production  had  indeed  met  increasing  competition  in  the  world's  market 
— from  Canada  and  Argentina,  in  particular — but  effects  were  always 
compensated  by  favorable  shifts  in  demand.  After  the  war,  this  was  no 
longer  so.  Although  population  increased  strongly,  consumption  per 
head  did  not.  On  the  contrary  the  latter  decreased  considerably  in 

1  We  are  speaking,  be  it  remembered,  of  the  pre-crisis  cotton  problem,  and  the  above  is 
intended  to  submit  that  even  before  and  independently  of  the  world  crisis  there  was  a  case 
for  what  we  call  planned  retreat.  It  should  be  clearly  understood  that  this  means  two 
different  things.  First,  locational  and  technological  innovation  in  the  ordinary  course  of 
the  cyclical  process  of  evolution  spelled  distress  for  a  great  part  of  the  Eastern  cotton- 
growing  industry.  In  a  competitive  industry  dominated  by  types  which  do  not  in  ration- 
ality and  promptitude  of  action  answer  to  capitalistic  requirements,  it  is,  of  course,  possible 
to  argue  for  planning  or  relief  in  order  to  facilitate  transitions,  which  in  themselves  are  normal 
and  unavoidable  but  which  the  people  themselves  are  unable  to  effect,  and  to  mitigate  the 
hardships  of  the  "competing-down  process."  But  there  was,  in  the  second  place,  another 
case  for  planning,  viz.,  for  planned  reaction  to  planned  expansion  abroad.  It  is  the  latter 
that  constitutes  the  special  feature  in  the  postwar  cotton  situation.  Whether  this  country 
was  to  enter  into  competition  with  the  new  sources  of  supply  opened  up  abroad  or  to  retire 
from  the  international  market  is  not  under  discussion  here;  either  course  would  have 
involved  planning  by  public  authority  in  order  to  be  successful.  Into  the  implications  of 
this  for  a  theory  of  planning  we  cannot  enter. 


738  BUSINESS  CYCLES 

response  to  changing  tastes  and  habits,1  though  the  increase  in  the  former 
was  sufficient  to  increase  total  consumption,  which  at  the  end  of  our 
period  was  about  15  per  cent  above  the  average  of  the  last  five  prewar 
years.  Foreign  demand  rapidly  fell  from  its  war  peak  after  the  cessation 
of  American  war  and  emergency  credits,  both  foreign  competition  and 
protection  accounting  for  the  sharply  falling  "trend"  and,  from  1926  on, 
the  uninterrupted  fall  in  quantity  exported.  World  production,  exclud- 
ing Russia  and  China,  after  having  decreased  from  1915  to  1917^,  increased 
to  1928  by  more  than  one-third  and  then  moved  about  a  level  approxi- 
mately 20  per  cent  above  the  last  prewar  years.2  European  production 
alone,  including  Russian  exports,  more  than  recovered  in  strongly  inverse 
covariation  with  United  States  exports.  Interpretation  of  these  facts 
must,  moreover,  take  into  account  the  very  low  elasticity  of  domestic 
demand.  Some  economists  hold  that  production  adapted  itself  to  the 
new  conditions  and  point  to  the  sharp  decrease  in  acreage  harvested  per 
head  of  population  that  occurred  from  1919  to  1925.3  Production  had, 
however,  expanded  considerably  from  1915  to  1919  (the  trough  of  1917 
was  due  to  the  failure  of  winter  wheat)  and  though  contraction  followed 
fairly  promptly  upon  the  fall  in  prices,  persisting  excess  capacity  and  that 
inelasticity  of  domestic  demand  would  nevertheless  account  for  strong 
effects  of  the  fall  in  exports  and  of  the  variations  in  harvests.4  It  does 
hot  follow  that,  because  the  price  of  wheat  moved  very  much  as  any 
all-commodity  index,  conditions  peculiar  to  the  wheat-growing  industry 
had  nothing  to  do  with  it.  The  argument  for  planned  retreat  thus  again 
suggests  itself. 

But  the  central  fact  was  the  technological  revolution.6  The  average 
yearly  product  during  our  decade  of  roughly  850  million  bushels  of  wheat 
may  not  look  formidable  in  itself.  But  it  was  not  the  result  of  harmon- 

1  But  Domestic  Disappearance  of  Flour,  as  calculated  by  the  Stanford  Food  Research 
Institute  (see  Wheat  Studies  for  December  1932,  p.  310),  fully  kept  pace  with  the  postwar 
percentage  of  population  growth.     For  grains  in  general  the  gradual  elimination  of  horses 
and  mules  must  also  be  taken  into  account. 

2  For  Argentina,  Australia,  and  Canada,  taken  together,  increase  as  compared  with  the 
last  prewar  years  was  over  40  per  cent.     Excepting  Argentina,  there  was  not  much  increase 
in  average  yield  per  acre. 

3  See,  for  example,  Warren  and  Pearson,  Prices,  Chart  42  on  page  50. 

4  Only  for  1918,  1919,  and  1924  did  good  harvests  spell  good  business  for  farmers.     The 
poor  crops  of  1920  and  1921  helped  to  steady  prices,  and  the  decade's  minimum  crop  (1925) 
brought  what  may  be  termed  recovery  values.     But  the  increasingly  good  harvests  of  1926, 
1927,  and  1928  exerted  a  pressure  unknown  to  prewar  times  and  led  up  to  the  collapse  of 
1929. 

5  This  has,  of  course,  never  been  entirely  overlooked.     But  the  paramount  importance 
of  that  element  of  the  case  has  been  first  pointed  out  by  E.  Altschul  and  F.  Strauss,  Tech- 
nological Progress  and  Agricultural  Depression,  National  Bureau  of  Economic  Research, 
Bulletin  67,  1937. 


1919-1929  739 

ious  expansion  in  all  parts  of  the  country,  which  it  would  have  been 
possible  to  restrict  again  at  proportionate  and  moderate  sacrifice  for  every 
grower  of  wheat  or  which,  in  fact,  would  have  restricted  itself  without 
catastrophe  in  the  course  of  a  few  years  of  depression.  It  was  the  net 
result  of  spectacular  expansion  in  some  regions  and  painful  elimination 
in  others.  Expansion  was  general  up  to  1919,  even  the  East  and  South 
responding  to  war  prices.  But  the  really  significant  increase  in  acreage 
was  not.  That  was  confined  to  Montana,  Kansas,  Nebraska,  Texas, 
and  a  few  other  states  and,  obviously,  was  not  due  simply  to  war  con- 
ditions. Similarly,  decrease  in  wheat  acreage  from  1919  to  1925  was 
general  but  also  unequal,  hardly  any  decrease  occurring  in  Montana, 
for  example.  The  subsequent  expansion  to  1929  coincided  with  restric- 
tion in  the  South  and  East,  where  acreage  decreased  by  about  one-fifth 
for  the  decade.  Diagnosis  of  this  course  of  events  is  obvious.  Expan- 
sion was  in  the  Great  Plains,  where  the  mechanized  farm,  the  tractor  and 
the  combine  thresher  in  particular,  can  be  worked  to  full  advantage  and 
yield  acceptable  returns  at  a  price  of  60  cents  per  bushel  or  less.1  Con- 
traction was  enforced  where  those  innovations  were  not  profitable  and  a 
price  of  one  dollar  per  bushel  covers  cost  only  on  the  better  soils.  We 
recognize  all  the  features  embodied  in  our  model  and  especially  the 
"competing-down  process,"  passing  sentence  of  economic  death  on 
perhaps  half  of  all  wheat  farmers.  The  implications  of  this  do  not  con- 
cern us  here.2  But  it  must  be  added  that  this  component  of  the  postwar 
situation,  in  fact,  originated,  as  it  should,  in  the  preceding  Kondratieff 
prosperity.  The  great  per-cent  increase  in  Montana,  Kansas,  Nebraska, 
and  Texas  was  from  1900  to  1915.  We  may  even  go  so  far  as  to  say 
that  this  is  what  can  be  attributed  to  innovation  per  se,  while  the  further 
rise  to  1919 — roughly,  10  million  acres  in  the  Great  Plains — was  a  war 
effect.  Innovation  would  have  spread  and  taken  full  effect  in  the  down- 
grade, as  it  always  does.  A  depressive  situation  would,  hence,  have 
ensued  in  any  case.  But  war  prices  and  reaction  to  them  accentuated  it, 
which  is  all  that  prices  or  monetary  factors  have  had  to  do  with  it. 

b.  For  England  and  Wales,  total  value  of  marketed  products  of 
domestic  agriculture  was,  in  1925,  225  million  pounds,  less  than  one-third 
of  consumption.  Two-thirds  of  that  amount  was  value  of  animal  prod- 
ucts; less  than  one-fourth,  value  of  cereals  and  potatoes.  Total  acreage 
in  cultivation  had  not  increased  at  all  during  the  war — it  had  slightly 
fallen — and  moved  below  the  average  of  1871-1875  throughout  our 
decade.  Prices,  of  course,  behaved  much  as  they  did  in  the  United 

1  See  Mr.  C.  L.  Holmes'  report  on  Farm  Production  Costs  as  affected  by  Mechanical 
Farm  Equipment,  United  States  Department  of  Agriculture,  1931. 

2  That  state  of  things,  of  course,  accounts  for  the  low  average  of  farmers'  income  per 
head. 


740  BUSINESS  CYCLES 

States;  relative  prices  of  animal  products  recovered  strongly  from  1924 
on.1  Income  from  agriculture,  as  per  income  tax  statistics — which  are 
not  very  reliable  in  the  case  of  agriculture — rose  to  1920-1921  (54  million 
pounds)  fell  to  1922-1923  (20.6  millions)  and  was  then  fairly  stable  (at 
about  24  millions)  until  1929.  In  the  absence  of  a  dangerous  load  of 
debt,  no  untenable  condition  resulted.  Investment  that  would  not  have 
paid  was  at  a  low  level.  But  the  medium-sized  farm  bore  up  well  on  the 
whole.  It  is  obvious  that  no  significant  effect  on  the  national  situation 
can  have  been  exerted.  Neither  was  there  any  "crisis"  until  1929-1930. 
There  was,  of  course,  the  boom  in  the  subsidized  sugar-beet  production, 
quantity  harvested  increasing  from  56,000  tons  in  1922-1923  to  1,472,300 
tons  in  1927-1928. 

c.  In  Germany,  inflation  had  reduced  the  agricultural  debt  from 
17.5  billion  gold  marks  (1913)  to  about  2.7,  the  amount  eventually  result- 
ing from  the  revaluation  of  mortgages.2  But  it  quickly  mounted  up 
again.  By  June  30,  1926  the  debt  newly  incurred — mortgage  and  other, 
but  not  counting  current  debts  with  retailers  and  so  on — amounted  to 
3.7  and  by  Mar.  31,  1930,  to  T.66  billions.  If  to  the  latter  figure  we  add 
2  billions  for  unreported  floating  debts  and  the  revaluation  mortgages, 
we  arrive  for  the  end  of  our  period  at  something  like  12  or  13  billions, 
and  at  an  interest  charge  of  over  1  billion.  In  order  to  understand  this, 
we  must  bear  in  mind  the  extent  to  which  German  agriculture  and  its 
equipment  had  suffered  during  the  war.  Considerable  investment  was 
necessary  to  restore  it  to  normal  efficiency.  The  inducement  to  run  into 
debt  was  stronger  for  the  larger  estates,  particularly  in  the  East,  because 
they  suffered  more  from  higher  wages  and  taxes,3  because  they  had  more 
opportunity  for  trying  out  new  methods  involving  investment,  because 
their  owners'  standard  of  life  proved  less  adaptable  than  that  of  the 
peasantry,  and  because  it  was  primarily  to  them  that  expensive  loans 
were  offered  from  public  and  private  sources  with  a  readiness  that  proved 

1  There  was,  in  spite  of  some  increase  in  pasture,  no  significant  increase  in  livestock. 
But  dairy  products  and  poultry  did  increase;  82  per  cent  of  the  butter  consumed  and  more 
than  half  of  the  meat  consumed  in  1924  was  imported,  however. 

2  No  account  need  be  taken  of  the  Rentenmarkgrundschuld,  i.e.,  the  amount  of  2  billion 
marks  for  which  German  land  had  been  mortgaged  by  law  in  order  to  provide  a  guaranty 
for  the  Rentenmarks  issued  in  the  process  of  monetary  stabilization.     Agriculture  had  to 
pay  5  per  cent  interest  on  this,  but  there  was  really  no  capital  debt.     When  the  Rentenbank 
was  liquidated  (act  of  Aug.  30,  1924)  it  was  succeeded  (act  of  July  18,  1925)  by  the  Benten- 
bankkreditanstalt,  which  took  over  its  assets  and  served  as  an  intermediary  agency  for  the 
loans  to  agriculture  that  came  from  public  and  foreign  sources. 

3  Taxes  on  agricultural  land  rose  more  in  the  West  than  in  the  East,  viz.,  from  8.8  marks 
per  nectar  in  1912-1914  to  37.6  marks  per  hectar  in  1924-1926  in  the  West,  and  from  7  to 
25.3  marks  in  the  East.     But  net  returns  shaped  much  more  unfavorably  in  the  latter 
region  than  they  did  in  the  former. 


1919-1929  741 

fatal.  Over  78  per  cent  of  all  holdings  were  eventually  in  debt  in  the 
West  and  almost  90  per  cent  in  the  East. 

If  we  are  to  trust  the  figures  presented  in  the  official  Survey  of  German 
Economic  Conditions  (Wirtschaftsenquete)  for  294  medium-sized  and 
large  units,  net  returns  per  hectar,  which  were  93  marks  in  the  average  of 
1912-1914,  were  practically  wiped  out  in  the  average  of  1924-1926. 
Another  semiofficial  source1  gives  3  marks  for  1924—1925  and  18  for 
1925-1926.  These,  of  course,  were  the  very  worst  years  until  1930  to 
1932,  and  net  returns  recovered  somewhat  in  the  interval.  Nor  do  such 
figures  disprove  the  impression  that  a  large  part  of  the  peasants,  especially 
those  in  thickly  populated  sectors  of  the  country,  did  quite  well.  But 
it  seems  impossible  to  avoid  the  inference  that  in  many  cases  returns 
barely  covered,  or  even  failed  to  cover,  taxes  and  interest  throughout  the 
period.  As  a  result,  foreclosures  (as  measured  by  acreage,  not  by  num- 
ber) rose  above  prewar  level  by  1926  and  then  steadily  increased  to  a 
maximum  of  193 1.2  This  does  not  even  tell  the  whole  tale,  because 
creditors  may  often  have  been  prevented  from  foreclosing  by  the  knowl- 
edge that  bidders  would  not  be  forthcoming  and  because  there  were  very 
many  desperate  situations  which  yet  stopped  short  of  technical  insol- 
vency. Prices  of  agricultural  land  were,  it  is  true,  even  in  1929-1930 
above  prewar  level,3  but  only  for  holdings  below  20  hectar.  For  larger 
ones  they  were  below  that  level,  the  more  so,  the  larger  the  holding. 

Diagnosis,  though  entirely  different  from  that  of  the  American  and 
English  cases,  is  not  difficult.  The  technological  component  plays  only 
a  minor  part.  According  to  the  census  of  1925,  German  agriculture  was 
then  but  imperfectly  mechanized,  even  in  the  class  of  units  above  100  hec- 
tar. The  use  of  sowing,  planting,  and  harvesting  machinery  was,  of 
course,  fairly  widespread  among  the  latter,  but  only  a  minority  used 
steam  and  motor  ploughs,  for  example,  and  the  majority  of  peasants  did 
practically  without  any  modern  machinery  at  all.  Only  644,713  of  the 
5.1  million  units  used  electromotors  and  the  tractor  had  not  yet  invaded  to 
a  significant  extent  the  domain  of  the  horse,4  the  ox,  or  the  cow.  There 
was,  indeed,  great  progress  in  fertilizing.  Average  yield  per  hectar  was, 
nevertheless,  for  all  grains  lower  in  1924-1928  than  it  had  been  in  1911- 

1  See  Veroffentlichungen  des  Deutschen  Landwirtschaftsrates,  Heft  16.     The  data  refer 
to  the  income  statements  of  those  units  that  reported  to  the  Accounting  Office  (Buchstelle) 
of  that  Council  of  Agriculture,  according  to  a  set  schema. 

2  After  that,  foreclosures  were  drastically  reduced  by  various  government  measures. 

3  See  W.  Rothkegel,  Die  Entwicklung  der  Kauf-und  Pachtpreise  fUr  LandgUter  und 
StUckl'andereien  in  the  publication  of  the  FriedrichList  Gesellschaft  on  Deutsche  Agrarpolitik 
I. 

4  There  were  in  Germany  (postwar  territory)  8.8  million  horses  in  1913  and  3.7  in  1928. 
The  use  of  horses  in  agriculture,  hence,  must  have  increased,  if  anything. 


742  BUSINESS  CYCLES 

1913,  and  land  in  agricultural  use  had  decreased.1  Germany's  dogged 
effort  toward  self-sufficiency  was  to  come  later.  It  must  not  be  for- 
gotten, however,  that  our  comparison  is  with  the  results  of  the  innova- 
tions of  the  Kondratieff  prosperity,  during  which  yield  per  hectar  had 
greatly  increased,  and  that  we  deal  with  national  averages  only,  which 
veil  significant  differences  between  different  regions.  But  such  further 
innovation  or  spread  of  previous  innovation  as  there  was  in  our  period 
hardly  affected  noninnovating  sectors  as  it  did  in  this  country,  because 
the  dominant  factor  of  the  German  situation  was  foreign  competition: 
all  domestic  innovation  did,  under  the  circumstances,  was  to  ease  things 
for  some  regions  or  individual  holdings. 

For  the  same  reason,  the  changes  in  consumers'  habits  mattered  less 
than  one  might  suppose.  But  while  demand  for  meat,  dairy  products, 
fruit,  and  vegetables  developed  satisfactorily  from  the  beginning  of  the 
period,  there  was  a  similar — though  less  pronounced — reduction  in  the 
consumption  of  breadstuff s  per  head,  as  there  was  in  the  United  States, 
coupled  with  a  shift  from  rye  to  wheat,  which  accounts  for  the  (small) 
increase  in  wheat  and  the  (greater)  decrease  in  rye  acreage,2  the  latter 
indicating  the  depressed  conditions  which,  with  the  exception  of  1927  and 
1928,  prevailed  in  the  most  important  branch  of  Germany's  production 
of  grains.  Those  effects  of  a  rising  standard  of  life  must  also  have 
influenced  potato  growing.  Conditions  varied  greatly  in  different  parts 
of  the  country,  and  the  statistics  of  direct  consumption  and  of  the  use 
of  potatoes  in  hog  and  cattle  rearing  and  as  an  industrial  raw  material 
present  difficulties  into  which  we  cannot  enter.  The  general  picture  is, 
however,  perfectly  clear.  A  technologically  very  rigid  supply,  which  had 
developed  in  prewar  times,  proved  too  much  for  the  market  of  a  com- 
modity that  was  in  the  process  of  losing  its  position  as  a  staple  food. 
The  ensuing  "surplus" — aided  by  a  similar  surplus  of  rye — partly 
explains  the  rapidity  with  which  the  hog  population  recovered  from  the 
massacre  inflicted  upon  it  as  a  war  measure:  in  1927  it  surpassed  its  pre- 
war figure.  This  and  competition  from  the  East  and  Southeast  brought 
down  the  price  of  pork,  so  that  all  the  most  important  products  of  Ger- 

1  From  1918  (postwar  territory)  to  1927  by  about  300,000  hectars.     But  the  figure 
should  not  be  implicitly  trusted.     That  decrease  may  be  due  to  difference  in  methods  of 
surveying.     There  was  certainly  no  significant  increase,  however.     Throughout  this  and 
the  following  chapter,  the  term  "postwar  territory"  refers,  unless  otherwise  stated,  to  the 
German  territory  as  delimited  by  the  Peace  of  Versailles. 

2  The  rye  acreage  fell  from  5.2  million  hectars  in  the  average  of  1911-1918  (postwar 
territory)  to  4.7  millions  in  the  average  of  1926-1928,  while  the  wheat  acreage  rose  from 
1.66  to  1.69  millions.     Conditions  in  wheat  growing  were  relatively  favorable  throughout 
the  period.     Value  of  the  harvest  surpassed  the  average  of  1908-1909  to  1913-1914,  which 
was  721.5  million  marks,  by  1925-1926,  when  it  was  791.2  millions.     Cf.  H.  Paetzmann, 
Zur  Lage  der  Deutschen  Landwirtschaft,  Vierteljahrshefte  zur  Konjunkturforschung,  1926, 
Sonderheft  No.  3,  p.  18. 


1919-1929  743 

man  agriculture  eventually  failed,  except  locally,  to  cover  costs  and 
burdens. 

Thus,  we  have  before  us  the  case  of  an  industry  which,  if  left  to  itself 
and  other  things  being  equal,  would  have  had  to  contract  ever  since  the 
late  nineties.  Protection  and  successful  innovation  enabled  it  to  hold 
its  own  and  even  to  expand  during  the  Kondratieff  prosperity.  But  in  the 
downgrade  the  superiority  of  American  opportunities — giving  much 
larger  scope,  in  particular,  to  mechanized  farming — and  the  cheap  labor 
of  the  countries  to  the  East  and  Southeast  of  Germany  were  bound  to 
assert  themselves.  It  then  became  clear  that  her  agriculture  was  unequal 
to  servicing  its  rapidly  growing  debt  and  to  the  wage  rates  and  burdens 
of  the  modern  state — that  its  marginal  value-productivity  was  hopelessly 
below  the  general  national  level.  It  is  from  this  angle  that  the  plight  of 
German  agriculture  must  be  understood.  The  state  of  things  could  not 
be  inferred  from  the  behavior  of  prices,  which  were  favorable  in  some 
cases,  especially  in  those  of  milk  and  butter,  and  not  obviously  catas- 
trophic in  any  case.  The  upsetting  factors  were  on  the  side  of  costs. 
Recognizing  this  yet  wishing  to  avoid  the  laisser-faire  conclusion  from 
this,  public  policy  exhausted  the  arsenal  of  protection,  prohibition  (at 
first  in  the  form  of  trade  monopolies),  reduction  of  railroad  rates,  direct 
and  indirect  subsidies,  and  also  resorted  to  "internal  colonization," 
breaking  up  a  number  of  larger  estates  into  small  settlements.  These 
measures  and  further  borrowing  kept  things  going  until  both  gave  way 
before  the  onset  of  the  world  crisis. 

2.  Next,  the  postwar  building  booms  call  for  comment.  Of  their 
quantitative  importance  in  the  economic  processes  of  the  period  it  is  very 
difficult  to  give  an  exact  idea,  but  very  easy  to  give  one  that  is  approxi- 
mate. If,  for  example,  we  accept  the  statement  contained  in  the  1929 
Census  of  Construction,  that  building  will  on  the  average  (directly) 
give  a  year's  full  employment  to  one  man  for  roughly  each  $5,800  spent, 
and  if  we  take  into  account  the  employment  created  by  the  production 
and  transportation  of  building  material  (which  we  might  now  attempt 
to  do  from  the  data  of  the  Federal  Employment  Stabilization  Board  and 
the  Bureau  of  the  Census)  and  in  other  subsidiary  industries  and,  by 
way  of  secondary  effects,  in  all  industries,  we  cannot  doubt  that  con- 
struction was  the  chief  contributor1  to  the  postwar  business  volume  in 
this  country  as  well  as  in  the  other  two.  This  is  no  more  unexpected  than 

1  As  will  presently  become  clear,  this  does  not  mean  that  there  was  a  one-way  relation 
between  building  and  the  rest  of  the  economic  organism,  or  that  the  prosperities  of  the 
period  originated  in  the  building  trade.  No  mere  appeal  to  the  quantitative  importance 
of  construction  within  the  total  of  system  expenditure  would  have  any  explanatory  value. 
In  most  cases  it  is  obvious  that  construction  rather  responded  to  than  created  conditions 
favorable  to  its  expansion.  The  latter  was  the  case  only  insofar  as  there  was  innovation 
in  the  building  industry  itself. 


744  BUSINESS  CYCLES 

the  postwar  agrarian  depression.  Building  booms,  in  particular  booms 
in  residential,  public,  and  public  utility  construction,  occurred  in  the 
downgrades  of  both  preceding  Kondratieffs — for  instance,  in  England 
in  the  twenties  of  the  nineteenth  century,  in  all  three  countries  before 
1873,  in  the  United  States  from  1878  to  1894.  All  of  them,  with  one 
exception,1  were  stronger  than  any  that  occurred  during  Kondratieff 
prosperities. 

Nor  is  this  a  mere  matter  of  history.  Taking,  for  brevity's  sake, 
dwelling-house  building  only,  we  need  but  list  the  factors  that  would 
produce  supernormal  activity,  in  order  to  see  that  the  general  conditions 
prevailing  in  Kondratieff  downgrades  and  revivals — more  precisely,  in  the 
prosperity  phases  of  the  shorter  cycles  which  run  their  course  within 
Kondratieff  downgrades  and  revivals — are  more  favorable  to  the  occur- 
rence of  building  booms  than  are  the  general  conditions  prevailing  in 
prosperities.  Falling  rate  of  interest  is  one  of  them.  High  rate  of 
increase  in  real  incomes  is  another:  from  rising  or  constant  money  incomes 
of  the  middle  and  lower  classes,  accompanied  by  falling  cost  of  living, 
new  demand  for  better  housing  will  naturally  follow.  Innovation  in  the 
building  industry  or  its  subsidiaries  will  work  in  the  same  direction 
because,  like  other  innovations,  it  is  likely  to  spread  in  recession.  The 
rise  in  rents  that  occurs  during  Kondratieff  prosperities  supplies,  barring 
a  subsequent  fall  in  money  incomes — which,  as  we  have  seen,  is  not 
likely  to  occur — an  additional  stimulus.  Finally,  industrial  evolution  in 
general  means  industrial  migration  and,  moreover,  migration  from  the 
countryside  to  the  cities,  both  of  which  create  new  demand  for  construc- 
tion that  is  eventually  provided  for  during  recession.  Of  course  there 
were,  besides,  other  factors,  unconnected  with  the  features  of  the  Kondra- 
tieff phase  which  happened  to  prevail.  The  omissions  of  the  war  period, 
both  as  to  replacement  and  as  to  normal  increment,  constitute  the  most 
important  of  them. 

a.  In  the  United  States  the  war  did  not  interfere  with  either  residential 
or  other  construction  to  anything  like  the  same  extent  as  in  England  or 
Germany,  but  such  indications  as  we  have  leave  no  doubt  about  the  fact 
that  it  was  at  an  abnormally  low  level  in  1917  and  1918.2  At  least 
in  most  parts  of  the  country,  a  short-lived  boom  set  in  at  the  end  of  1918, 
during  which  building  costs  rose  sharply — by  25  per  cent  or  more.  This 
was  followed  by  a  drastic  fall  both  in  building  activity  and  costs,  and 
from  the  beginning  of  the  fourth  quarter  of  1923  the  postwar  boom  in 
residential  building  definitely  got  under  sail.  There  was  a  setback  in  the 

1  That  exception  is  the  English  building  boom  from  1895  to  1905.     It  set  in  during 
revival,  but  it  persisted  through  nearly  half  of  the  subsequent  Kondratieff  prosperity. 

2  Cf.  J.  R.  Riggleman,  Building  Cycles  in  the  United  States,  1875-1932,  Journal  of  the 
American  Statistical  Association^  June  1933. 


1919-1929  745 

second  and  third  quarters  of  1924;  then  a  peak  was  reached  in  1925, 
descent  from  which  lasted  to  about  May  1927;  another  peak  occurred  in 
April  1928.  After  that  we  have  decline,  which  though  at  varying  rates — 
in  1930  there  was  some  retardation — continued  to  February  1933,  with 
1929  and  1931  displaying  the  sharpest  falls.  In  apartment -house  and 
hotel  construction  the  maximum  occurs  in  1926,  but  the  figure  for  1925 
comes  near  it  and  those  for  1927  and  1928  are  not  much  below  it.1 
Expenditure  on  new,  nonfarm,  residential  building,  including  hotels  and 
clubs,  is  estimated  by  the  National  Bureau  of  Economic  Research  at 
34  billion  dollars  for  the  decade. 

Diagnosis  of  that  boom,  which  was  entirely  financed  from  private 
sources,  presents  no  difficulties.  At  the  beginning  of  the  period  there  was 
dammed-up  demand.  Population,  in  spite  of  the  Immigration  Restric- 
tion Act  of  May  1921,  increased  from  1920  to  1929  by  15  millions,  the 
largest  absolute  amount  of  increase  per  year  in  the  history  of  the  country. 

1  On  the  sources  of  construction  statistics  and  some  of  the  pitfalls  they  contain  see  C. 
Gill,  Construction  Statistics,  Journal  of  the  American  Statistical  Association,  March  1933. 
The  above  refers  to  values  of  contracts  awarded  and  is  mainly  based  on  the  data  of  the  F. 
W.  Dodge  Corporation,  which  until  1932  exclude  contracts  of  a  value  less  than  $5,000  and 
— an  advantage  from  our  standpoint — alterations  and  repairs.  The  data,  steadily  increas- 
ing in  accuracy,  specialization,  and  coverage,  start  in  1901,  cover  27  states  from  1922,  36 
from  1923,  and  later  on  37,  from  which  the  national  total  is  arrived  at  by  a  multiplier.  The 
writer  wishes  to  acknowledge  the  F.  W.  Dodge  Corporation's  courtesy  in  answering  his 
questions  and  permitting  use  of  material.  No  other  series  of  contracts  awarded  have  been 
used  except  to  test  the  writer's  guesses  as  to  particular  questions.  Building  permits  are 
available  (Carl  Snyder's  series)  for  seven  cities  since  1882.  Of  other  permit  series,  the 
annual  one  published  by  the  Bureau  of  Labor  Statistics — on  account  of  its  specialized 
information — and  the  monthly  one  published  by  the  Federal  Reserve  Board  have  been  most 
useful.  But  a  report  by  the  National  Bureau  of  Economic  Research,  foreshadowing  the 
results  of  a  comprehensive  investigation  working  with  permits  (D.  L.  Wickens  and  R.  R. 
Foster,  Non-farm  Residential  Construction,  1920-1936,  Bulletin  65,  Sept.  15,  1937)  seems, 
on  the  one  hand,  entitled  to  so  much  confidence  and,  on  the  other  hand,  to  yield  results  so 
materially  different  that  the  writer  is  now  inclined  to  doubt  the1  reliability  of  his  own  work 
in  the  matter  summarized  in  the  text.  In  particular,  the  second  peak  (1928)  may  be 
untrustworthy.  Instead,  new  residential  construction  is  held  by  the  authors  of  that  report 
to  have  declined  steadily  from  the  first  peak  (1925).  But  there  is  no  difference  as  to  the 
abruptness  of  the  fall  in  1929  or  the  steepness  of  the  ascent  in  1925.  Physical  volume  (there 
are  now,  however,  new  estimates  of  numbers  of  units  constructed,  e.g.,  L.  J.  Chawner's  in 
the  Annals  of  the  American  Academy  of  Political  and  Social  Science,  vol.  190)  must  be  derived 
from  those  value  figures  by  deflating.  Several  indices  have  been  constructed  for  this 
purpose.  They  include,  however,  labor  and  materials  only.  The  one  used  by  the  writer, 
published  by  the  Federal  Reserve  Bank  of  New  York,  attributes  (constant)  weights  of 
respectively  45  and  55  per  cent  to  those  two  items  (weighting  according  to  the  1929  Census 
of  Construction:  labor  33  materials  47  per  cent,  other  items  20  per  cent).  Professor  Mills 
(Economic  Tendencies  in  the  United  States,  p.  267)  rightly  deflates  the  dollar  volume  of 
each  type  of  building  by  an  index  appropriate  to  it.  The  reader  is  referred  to  his  treatment 
of  the  subject.  But  from  1925  to  1929  all  cost  indices  change  so  little  that  the  general 
picture  is  affected  by  deflation  only  for  1919  to  1924. 


746  BUSINESS  CYCLES 

There  was  also  considerable  internal  migration.  Real  income  per  head, 
rising  strongly  in  all  brackets,  made  that  demand  effective  and  added 
new  sources.  The  motor  was  the  only  other  "expensive"  one  of  the 
items  toward  which  the  surplus  turned.  From  1916  to  1920  rents  had 
risen  on  a  national  average  by  almost  two- thirds.  They  fell  but  insig- 
nificantly, even  in  the  crisis  of  1921.  Primarily  the  boom  was  a  response 
to  these  conditions.  Building  costs  rose  swiftly  in  1923  and  after  1924 
remained  fairly  stable  on  a  somewhat  lower  level.1  Interest  on  urban 
mortgages  was,  though  falling,2  not  particularly  cheap  as  compared  with 
other  long-term  rates,  except  where  building  was  financed  by  bond  issues. 
But  under  the  circumstances  of  that  period  and  in  the  glow  of  its  uncritical 
optimism  neither  costs  nor  interest  charges  mattered  much.  It  seemed 
more  important  to  get  quickly  the  home  one  wanted — or  the  skyscraper 
the  prospective  rents  of  which  in  any  case  compared  favorably  with  the 
rate  on  mortgage  bonds — than  to  bother  whether  it  would  cost  a  few 
thousand  dollars — or  in  the  case  of  the  skyscraper,  a  million  or  so — more 
or  less,  provided  money  was  readily  forthcoming  at  those  rates.  And 
it  was.  First  mortgages  on  urban  real  estate  represent,  on  the  one  hand, 
not  all  the  loans  that  were  made  available  for  building  and,  on  the  other 
hand,  also  financed  not  only  other  types  of  building  but  other  things  than 
building.  But  it  is  still  permissible  to  point  to  the  fact  that  they  increased 
from,  roughly,  13  billions  in  1922  to,  roughly,  27  in  1929,  building  and 
loan  associations  contributing  about  7.8,  commercial  banks  about  5.2, 
mutual  saving  banks  5.1,  life  insurance  companies  4.8,  and  mortgage 
bonds  more  than  4.3  This  increase  is  out  of  all  proportion,  not  only  with 
the  increase  in  what  can  in  any  reasonable  sense  be  called  savings,  but 
also  with  the  expansion  of  bank  credit  in  other  lines  of  business,  and 
illustrates  well  how  a  cheap  money  policy  may  affect  other  sectors  than 
those  in  which  it  is  conspicuously  successful  in  bringing  down  rates. 

1  The  increase  in  building  costs  in  1923  is  beyond  doubt.     The  decrease  that  occurred 
in  1924  was  a  few  per  cent  only  and  figures  out  differently  with  different  indices.     No 
substantial  reduction  occurred  before  1931,  at  least  as  far  as  quoted  prices  are  concerned. 

2  It  is  interesting  to  note  that  the  beginning  of  the  boom — like  that  of  the  English  boom 
of  1982  (see  infra,  Chap.  XV,  Sec.  D) — preceded  such  fall  as  there  was  in  mortgage  rates. 
In  this  as  in  other  cases  it  is  perfectly  clear  that  other  factors  are  more  important  than 
interest,  although  our  description  sufficiently  shows  that  the  writer  has  no  intention  to  deny 
its  influence.     Residential  building  fell  off,  however,  at  the  very  time  bond  yield  began 
to  fall  more  markedly. 

8  Those  figures  are  taken  from  C.  E.  Persons,  Credit  Expansion,  1920-1929,  and  its 
Lessons,  Quarterly  Journal  of  Economics,  November  1930.  They  differ  somewhat  from 
those  of  J.  H.  Gray  and  G.  W.  Terborgh,  First  Mortgages  in  Urban  Real  Estate,  Report 
for  the  Real  Estate  Research  Committee  of  the  Brookings  Institution,  1929.  Both  esti- 
mates exclude  second  mortgages  and  other  items.  Professor  Irving  Fisher,  Booms  and 
Depressions,  1932,  p.  173,  puts  the  total  nonfarm  mortgage  debt  in  1929  as  high  as  37  billion 
dollars(?). 


1919-1929  747 

If  such  a  sector  display  a  very  elastic  demand  for  the  funds  which  that 
policy  will  drive  toward  it,  interest  in  it  need  fall  but  little  or  not  at  all  in 
order  to  produce  all  the  consequences  that  we  usually  associate  with 
"too-low"  money  rates. 

Innovation  lent  its  aid.  The  steel-skeleton  structure,  made  cheaper 
by  steadily  increasing  use  of  reinforced  concrete  and  workable  by  the 
electric  elevator,  had  created  new  possibilities  ever  since  the  nineties,  and 
these  possibilities  had  become  realities — and  a  feature  of — the  Kondra- 
tieff  upswing.  In  the  downgrade  after  the  war  this  innovation,  improved 
by  several  minor  and  "induced"  ones,  propelled  by  changes  in  the  habits 
of  life  that  made  the  apartment  increasingly  desirable  to  the  American 
bourgeois  family  and  by  the  plethora  of  credit,  spread  and  conquered, 
much  like  motorcars  or  rayon  and  exactly  like  those  innovations  that 
carried  the  prosperities,  and  spread  in  the  downgrades,  of  preceding 
Kondratieffs.  Similarly,  prefabrication,  primarily  made  possible  by 
the  use  of  the  new  materials  but  also  applied  to  stone  and  lumber, 
extended  its  domain  far  beyond  the  skyscraper.  Excavation  of  base- 
ments by  means  of  power  shovels  improved  by  the  caterpillar  tread  and 
belt  and  bucket  conveyors,  the  use  of  power  hoists,  of  power  concrete 
and  mortar  mixers,  and  of  pneumatic  riveting  machines,  rapidly  became 
a  matter  of  course  for  contractors  in  all  lines  of  building1 — typical  down- 
grade developments,  all  of  them.  Their  full  effect — the  mass  production 
of  the  perfectly  standardized  and  mechanized  cheap  house — is  still  to 
come,  however.  During  our  period  the  ordinary  family  house  was  in 
the  main  still  being  built  in  substantially  old-fashioned  ways  by  small 
and  inefficient  firms. 

But  the  conclusion  that  this  essentially  consequential  development — 
in  response  to  the  omissions  of  the  war  period  and  increased  real  pur- 
chasing power  of  all  classes,  on  the  one  hand,  and  in  response  to  previous 
innovation,  on  the  other — issued  in  overbuilding,  owing  to  the  additional 
stimulus  imparted  by  the  monetary  factor,  must  not  be  accepted  hastily, 
however  plausible  it  may  seem.  Some  types  of  response  to  those  con- 
ditions, especially  the  ones  that  were  linked  to  speculative  real  estate 
operations,  were  clearly  of  the  bubble  class.  The  Miami  case  may  serve 
as  an  example.  Nor  can  there  be  any  doubt  about  the  merits  of  the 
financial  methods  that  were  also  used  in  less  "speculative"  cases — New 
York  skyscrapers  for  instance — and  in  particular  about  the  financial 
quality  of  the  mortgage  bonds,  which  increased  from  682  millions  in  1922 
to  4,169  millions  in  1929,  and  which  were  readily  lent  against  by  banks. 
Finally,  everything  was  done  to  make  it  easy  for  everyone  to  run  into 
debt,  for  the  purpose  of  building  a  home  as  for  any  other  purpose.  It  is 
indeed,  easy  to  understand  that  such  a  structure  would  give  way,  not 
1  See  Jerome,  Mechanization  in  Industry,  pp.  134-145. 


748  BUSINESS  CYCLES 

only  under  the  impact  of  a  serious  crisis,  but  even  in  consequence  of  a 
mere  failure  of  rosy  expectations  about  things  in  general  to  come  true. 
In  other  words,  we  shall  readily  understand  why  the  load  of  debt  thus 
lightheartedly  incurred  by  people  who  foresaw  nothing  but  booms 
should  become  a  serious  matter  whenever  incomes  fell,  and  that  con- 
struction would  then  contribute,  directly  and  through  the  effects  on  the 
credit  structure  of  impaired  values  of  real  estate,  as  much  to  a  depression 
as  it  had  contributed  to  the  preceding  booms.  Nothing  is  so  likely  to 
produce  cumulative  depressive  processes  as  such  commitments  of  a  vast 
number  of  households  to  an  overhead  financed  to  a  great  extent  by  com- 
mercial banks.  But  this  does  not  quite  amount  to  saying  that  there  was 
overbuilding  in  the  sense  that  the  amount  of  construction  was  greater 
than  it  was  possible  to  absorb  without  losses  under  the  conditions  then 
prevailing,  and  that  this  excess  was  an  independent  cause  of  the  great 
depression. 

Rents  fell  from  1924  on,1  but  only  moderately.  Vacancies  increased, 
but  not  more  than  was  to  be  expected  in  a  period  of  rapid  obsolescence  of 
existing  house  property.  The  big,  old,  ugly,  and  inconvenient  house 
soon  became  difficult  to  sell,  because  of  changes  in  tastes — some  of  them 
attributable  to  the  automobile — and  because  of  the  increasing  wages 
and  decreasing  efficiency  of  servants.  But  there  is  no  reason  to  believe — 
in  particular,  the  fact  that  the  increase  in  dwelling  units  was  greater 
than  the  net  increase  in  married  couples  does  not  prove — that  the  spurt  of 
1925  could  not  have  settled  down  into  an  appropriate  average  activity 
and  that  even  the  results  of  speculative  excesses  could  not  have  been 
liquidated  without  any  violent  crisis  in  building,  let  alone  in  general 
business.  As  a  matter  of  fact,  this  was  accomplished  to  a  certain  extent. 
If  we  accept  the  figures  of  the  National  Bureau  of  Economic  Research, 
we  arrive  at  the  conclusion  that  four  years  of  such  adjustment — including 
local  crises — actually  followed  upon  that  boom  without  much  general 
disturbance  being  created.  In  the  final  result,  expansion  in  this  line  was 
not  so  obviously  greater  than  it  was  in  other  lines  of  consumption  that 
explanation  of  subsequent  vicissitudes  could  simply  be  given  in  terms  of 
"malin  vestment."  And  incomes  had  first  to  fall  because  of  a  general 
crisis,  for  the  special  crisis  of  building  and  of  real  estate  to  come  about.2 

1  All  indices  agree  in  this;  there  is,  however,  some  doubt  whether  or  not  the  national 
average  temporarily  increased  again  somewhat  in  1926  and  1927. 

*  The  reader  is  welcome  to  argue,  if  he  feel  so  inclined,  that  the  above  comes  to  saying 
that  all  that  would  have  been  necessary  in  order  to  avert  serious  trouble  would  have  been 
continued  increase  of  incomes  by  means  of,  say,  additional  government  expenditure.  It 
is  more  important,  as  well  as  more  useful,  to  point  out  that  timely  restriction  was  not  the 
only  alternative  open  under  the  circumstances.  Much  could  have  been  achieved  by  devis- 
ing more  appropriate  methods  of  financing  by  public  or  semipublic  institutions,  which 
might  have  kept  commercial  banks  out  of  this  business  and  created  new  forms  of  loan 
contracts  less  sensitive  to  the  impact  of  depressions. 


1919-1929  749 

This  analysis  refers  to  residential  building  only.  Results  are  not, 
however,  substantially  changed  by  including  other  types.  One  of  them, 
commercial  building,  is  perhaps  still  more  than  apartment  houses  and 
hotels  exposed  to  the  suspicion  of  speculative  overdoing.  Contracts 
awarded1  increased  steadily  to  a  peak  in  1927  and  another  almost  as  high 
in  1929,  and  summed  up  for  1922  to  1929  to  nearly  6.7  billions,  the  rate  of 
increase  over  the  period  being  substantially  in  excess  of  that  of  residential 
building.  Industrial  building  increased  at  a  still  greater  rate — contracts 
awarded  sum  up  to  about  4.8  billion — but  there  is  very  little  reason  to 
suspect  any  excess  over  the  requirements  of  the  general  march  of  things. 
Unlike  the  other  items,  but  also  conforming  to  expectation,  this  moved 
well  in  the  Kitchins  and  showed  equally  well  the  sweep  of  the  two  incom- 
plete Juglars.2  Finally,  more  than  one-third  of  the  grand  total3  of  con- 
tracts awarded — nearly  49  billions  according  to  the  Dodge  figures,  which 
certainly  understate — comes  under  the  heading  of  Public,  Institutional, 
and  Utility  construction.  Part  of  this  is  reflected  by  the  increase  in 
municipal  bonded  debts.  Alone  federal  expenditure  on  new  construc- 
tion, repairs,  and  alterations  amounted,  from  1920  to  1929,  according  to 
the  Federal  Employment  Stabilization  Board,  to  about  2.5  billions,  the 
trend  being  upward  all  the  time  and  1929  displaying  the  highest  figure 
(308  millions4) — a  fact  worth  mentioning  in  view  of  the  prevalent  talk 
about  insufficient  spending.  According  to  the  same  source,  the  figures 
of  which  are  again  incomparable  with  those  used  above,  the  expenditure 
of  railroads — steam  and  electric — and  power  and  telephone  companies 
on  construction  and  maintenance  moved,  from  1923  to  1929,  extremely 
steadily  on  a  slightly  rising  level,  summing  up  to  20.4  billions.  But  these 
sums,  of  the  importance  of  which  those  examples  suffice  to  give  an  idea, 
were  expended  in  ways  that  would  not  produce  any  material  effect  on  the 
economic  process  beyond  what  is  implied  in  the  expenditure  itself. 

b.  In  Germany  privately  financed  construction  during  the  war  was 
probably  at  less  than  one-fifth  of  the  1910  to  1913  average  and  in  the  end 
practically  ceased.  Some  impulse  was  given  to  it  by  the  "flight  into 

1  According  to  the  Dodge  figures  of  value  of  contracts  awarded.     They  must  not  be 
added  to  or  compared  with  the  National  Bureau's  estimates  of  residential  building.     For 
this  purpose  they  would  have  to  be  increased,  though  presumably  not  as  much  as  the  Dodge 
estimates  of  dollar  volume  of  residential  building. 

2  The  figure  for  1923  was  almost  80  per  cent  above  that  of  1922;  1924  shows  decrease; 
1925,  over  80  per  cent  above  1924,  indicates  the  rise  of  the  fourth  Juglar,  which  starts  in 
the  second  half  of  that  year;  1926  reflects  prosperity  in  the  act  of  gathering  momentum; 
1927  is  influenced  by  the  Kitchin  depression;  1928  shows  recovery;  and  1929  brings  the 
peak  of  the  period  (845  millions). 

8  It  is  worth  while  noticing  that  that  total  which  reached  its  peak  in  1928  was  very 
stable  from  year  to  year,  1925  to  1928,  and  fell  by  12  per  cent  in  1929. 

4  This  expenditure  then  rose  to  new  levels  in  1930,  1931,  and  1932.  For  this  and  the 
following  statement,  see  C.  Gill,  op.  cit.,  pp.  39  and  40, 


750  BUSINESS  CYCLES 

real  values"  during  the  inflation  period,  but  this  impulse  ceased  to  act 
in  the  time  of  "wild"  inflation.  From  1924  to  1929  industrial  and  com- 
mercial building  presents  a  picture  not  unlike  the  American  one.  It  rose 
from  1.23  billion  marks  in  1925  to  2.16  in  1925,  fell  in  1926,  more  than 
recovered  in  1927,  and  reached  its  peak  of  nearly  3  billions  in  1928, 
when  commercial  and  industrial  contracts  awarded  in  the  United  States, 
not  yet  at  their  peak  however,  amounted  to  about  6.7  billion  marks. 
Though  the  latter  figure  understates  actual  expenditure  to  an  unknown 
extent,  it  is  nevertheless  clear  that,  considering  the  difference  in  the 
size  of  the  respective  economic  organisms,  Germany  did  in  this  respect 
not  much  worse  than  this  country.  There  was  a  small  fall  already  in 
1929,  but  for  1930,  the  further  fall  being  but  11  per  cent,  the  comparison 
is  still'more  favorable.  The  total  from  1924  to  1929  was  13.45  billion 
marks,  about  1  billion  greater  than  the  total  for  public  building — includ- 
ing roads  and  canals — which  behaved  similarly  but  fell  off  much  more 
than  business  building  in  1930. 1  Residential  building  was  subsidized, 
as  has  been  stated  before.  It  is  doubtful  whether  it  would  have  revived 
of  itself  under  the  circumstances,  since  the  rigorous  control  of  rents  of  old 
dwellings  absorbed  part  of  the  demand  and  made  people  unwilling  to 
pay  rents  which  would  have  covered  costs.  As  it  was,  great  building 
activity  set  in,  expenditure  steadily  increasing  to  a  peak  in  1928 — 
followed  by  a  considerable  drop — of  3.4  billion  marks,  which  compares 
with  about  3  billion  dollars  in  the  American  peak  year,  1925,  or  with  4.7  if 
we  accept  the  National  Bureau  estimate:  the  comparison,  however 
precarious,  because  residential  building  does  not  cover  exactly  the  same 
things  in  both  countries  and  for  other  reasons,  is  yet  not  without  sig- 
nificance considering  the  difference  in  real  income  per  head.  The  grand 
total  for  the  period  was  40  billions,  to  which  residential  building  con- 
tributed 14  billions.  Unlike  the  corresponding  American  amount,  this 
sum  largely  failed,  independently  of  any  crisis,  to  give  economic  return.2 

1  The  madness  of  this  kind  of  reaction  to  the  world  crisis — for  madness  is  the  mildest 
term  the  writer  can  think  of  when  placing  himself  on  the  standpoint  of  the  gospel  of  spend- 
ing— ceases,  however,  to  be  so  obvious  when  the  internal  and  external  political  situation 
and  the  previous  fiscal  policy  are  properly  taken  into  account,  see  the  following  chapter. 
We  should  note  here  that,  especially  in  the  last  years  of  the  period,  the  financing,  par- 
ticularly of  municipal  building,  was  the  last  word  in  "  unsoundness  "  and  may  compete  with 
anything  skyscraper  financiers  did  in  this  country.     In  some  cases  such  things  as  bridges 
were  financed  by  six  months'  paper.     The  total  may  also  have  been  understated,  for  some 
items  were  tucked  away  in  municipal  budgets  of  extraordinary  reticence. 

2  The  Survey  of  Economic  Conditions  (Ausschuss  zur  Untersuchung  der  Erzeugungs 
und  Absatzbedingungen  der  deutschen  Wirtschart.     Third  subcommittee,  Der  Deutsche 
Wohnungsbau,  1931,  p.  18)  records  the  opinion  that  economic  calculation  was  put  out  of 
operation  (ausser  Kraft  gesetzt)  in  that  field.     Although  intended  to  be  cautiously  reserved 
and  perhaps  the  result  of  a  compromise,  the  passage  is  misleading  and  even  unjust.     As 
pointed  out,  there  was  reckless  financing  and  there  may  have  been  some  waste.     But 


1919-1929  751 

c.  It  was  in  August  1932  that  the  great  English  building  boom  began 
that  was  to  solve  the  housing  problem  in  the  same  sense  in  which  it  can  be 
said  that  the  development  of  the  textile  industries  in  the  three  decades 
following  the  Napoleonic  wars  solved  the  clothing  problem  and  the 
development  of  agriculture  after  the  World  War  solved  the  food  problem 
of  the  masses — leaving  many  things  to  be  done,  no  doubt,  but  only  things 
of  the  second  order  of  difficulty  and  importance.1  That  was  the  housing 
boom  which  corresponds  to  the  American  one  in  the  twenties.  The 
building  activity  during  the  period  under  discussion  presents  essentially 
different  features.  Our  remarks  will,  however,  be  confined  to  residential 
building  also  in  this  case. 

The  general  conditions  of  a  Kondratieff  downgrade  would,  ceteris 
paribus,  have  been  as  favorable  in  England  as  they  were  elsewhere  to  the 
occurrence  of  a  housing  boom.  In  1912  and  1913  such  a  boom  actually 
seemed  about  to  develop  in  continuation  of  the  one  that  started  in  the 
nineties.  And  after  the  war  there  was,  of  course,  dammed-up  demand; 
the  number  of  marriages  rose  considerably,  if  temporarily,  above  the 
1913  figure  and  so  on.  But  other  things  were  not  equal,  and  one  feature 
of  Kondratieff  downgrades  failed  for  the  time  being  to  materialize: 
at  least  to  1924,  per  capita  real  income  was  not  much  higher2  than  it 
had  been  10  years  earlier.  Hampered  also  by  high  costs,  building  there- 
fore showed  little  tendency  to  revive  in  1919.  While  in  this  respect  the 
English  was  dissimilar  to  the  American  case,  it  was  dissimilar  to  the 
German  case,  not  only  in  the  resources  that  were  available  to  cope  with 
the  situation,  but  also  in  the  sober  spirit  in  which  it  was  met.  This 
spirit  asserted  itself  in  three  ways.  First,  rent  control,  imposed  during 
and  continued  after  the  war,  was  administered  in  a  much  more  business- 
like way  than  in  Germany.  Second,  private  enterprise — especially 
large-scale  private  enterprise — was  harnessed  to  the  purpose  rather  than 
discouraged,  and  a  workable  mixture  of  public  and  private  activity  was 
experimentally  found.  One  of  the  contributions  of  this  policy  to  the 
boom  of  the  thirties  was  precisely  that  it  trained  the  building  industry  to 
its  task,  while  at  the  same  time  the  appetite  of  people  who  could  afford 
to  pay  for  better  housing  was  being  stimulated.  Third,  it  was  recognized 
in  fact — whatever  actual  motives  and  phraseologies  may  have  been — that 
the  road  toward  satisfactory  housing  of  the  lowest  strata  led  through 

extensive  dwelling-house  building  was,  under  the  general  conditions  of  the  time,  the  Kond- 
ratieff phase  included,  a  perfectly  normal  phenomenon,  and  a  public  contribution  toward 
the  creation  of  satisfactory  homes  need  not  have  created  financial  trouble. 

1  This  is  obvious  in  the  case  of  food  and  clothing,  and  for  housing  conclusively  proved 
by  the  Ministry  of  Health  report  on  overcrowding  in  England  and  Wales,  1986. 

2  According  to  Professor  Bowley,  it  was,  if  anything,  somewhat  smaller.     The  question 
will  be  discussed  in  Sec.  E. 


752  BUSINESS  CYCLES 

providing  houses  for  that  large  number  of  families  of  intermediate  posi- 
tion— skilled  workmen,  clerks,  those  in  the  lower  ranks  of  the  professions, 
and  so  on — whose  incomes  were  adequate,  or  nearly  so,  to  cover  costs, 
and  who  needed  but  little  direct  assistance  which  would  involve  only 
moderate  and  calculable  sacrifices  to  the  exchequer,  although  guarantees 
might  be  required  to  make  the  structure  of  the  necessary  credit  crisis- 
proof.  Slum  clearance  would  then  do  the  rest. 

This  policy  was  adopted  by  the  Chamberlain  Act  (1923  to  1929; 
amended  by  the  Financial  Provisions  Act  of  1924),  which  provided  the 
backbone  of  the  house-building  activity  during  our  period.  The  Addison 
Act  (1919,  restrictively  amended  1921)  had  started  the  movement  by 
vesting  the  task  of  building  houses  with  local  authorities  and  guaranteeing 
them  against  loss.  The  Wheatley  Act  (1924)  again  aimed  chiefly  in  the 
direction  of  building  by  local  authorities.  But  the  Chamberlain  sub- 
sidies really  did  the  job.  The  effect  of  these  measures,  which  eventually 
burdened  the  exchequer  with  not  more  than  about  13  million  pounds  a 
year — a  burden  that  was,  moreover,  compensated,  at  least  in  part,  by  the 
increase  in  revenue  and  (relative)  decrease  in  unemployment  expenditure 
which  they  at  the  time  certainly  entailed1 — was  to  raise  the  annual  number 
of  units  built  above  the  annual  average  of  the  last  prewar  decade  by 
19242  and  to  a  maximum  of  273,230  units  in  1926-1927,  excluding  houses 
above  78  pounds  or,  in  the  metropolitan  area,  105  pounds  ratable  value. 
In  all — counting  in  the  tail  end  in  the  thirties,  consisting  of  about  257,000 
units  built  under  the  Wheatley  Act — 1,177,863  subsidized  houses  were 
erected  at  an  estimated  cost  of  671  million  pounds,  local  authorities  being 
responsible  for  756,298  houses  and  419  millions,  private  enterprise  for  the 
rest.  These  subsidies  tapered  off  from  1928  on  and  were,  pro  futuro, 
abolished  by  the  Housing  Act  of  1933,  when  this  policy  had  been  aban- 
doned in  favor  of  direct  attack  on  the  problem  of  the  lowest  strata  by 
means  of  slum  clearing  (Act  of  1930).  By  then,  however,  conditions  had 
begun  to  shape  for  the  expansion  of  unassisted  building,  which  increased 
in  1928,  when  subsidized  construction  fell  off  sharply — there  was  in  the 
total  a  net  decrease,  of  course — continued  to  increase  through  1931,  and, 
after  falling  but  insignificantly  in  1932,  embarked  upon  its  astounding 
performance. 

No  problem  arises,  however,  for  our  period.  The,  roughly,  71,000 
houses  built  without  assistance  in  1928—1929  still  compare  with,  roughly, 
133,000  subsidized  units  and  are,  if  anything,  less  than  could  have  been 

1  The  theory  of  this  is  due,  as  the  reader  knows,  to  Mr.  R.  F.  Kahn.     In  its  bearings 
upon  our  case,  the  theory  of  Secondary  Employment  is  not  invalidated  by  Professor 
Neisser's  criticism  in  Review  of  Economic  Statistics  for  February  1936. 

2  That  prewar  average  was  about  100,000.     In  1922  over  106,000  houses  were  built 
under  the  Addison  Act,  but  this  was  due  to  an  effort  to  profit  from  the  act  before  the 
restrictions  of  1921  became  effective. 


1919-1929  753 

expected  from  tlie  fall  in  costs — building  costs  only — to  little  more  than 
one-third  of  the  1920  figure.  Another  factor,  which  in  the  thirties  was 
to  play  so  great  a  role  was,  however,  gathering  power  all  the  time.  Build- 
ing Societies,  the  history  of  which  goes  back  to  the  last  decades  of  the 
eighteenth  century  and  which,  favored  by  public  policy  in  many  ways, 
had  made  considerable  headway  since  the  eighties  of  the  nineteenth, 
increased  their  assets — mostly  mortgages — which  in  1918  were  only  about 
5  per  cent  above  what  they  had  been  in  1913,  from  77.3  million  pounds  in 
1919  to  312.7  in  1929.  This  increase,  significant  also  for  other  purposes 
than  the  one  in  hand,  was  as  steady  as  it  was  great.  Together  with  their 
organization — ideally  adapted  to  the  financing  of  houses  in  the  price 
range  from  400  to  500  pounds — and  their  privileges,  this  made  them  the 
leaders  in  that  line  of  business.  Insurance  companies  were  their  chief 
competitors.  The  success  of  both  is  an  outstanding  symptom  of  the 
extent  of  the  transfer  of  wealth  that  had  occurred. 

It  is  obvious,  then,  that  during  that  period  English  residential  build- 
ing was  primarily  a  function  of  subsidies  and  of  expectation  concerning 
subsidies  but  not,  as  in  the  United  States,  of  cyclical  phases.  It  shaped 
business1  situations  rather  than  being  shaped  by  them  and  thus  "dis- 
turbed" the  course  of  cycles.  Considering  the  quantitative  importance 
of  this  influence  we  shall  hence  expect  what,  from  the  standpoint  of  our 
schema,  would  have  to  be  called  irregularities.  As  it  happens,  they  are 
not  very  great.  Needless  to  say  that,  even  if  they  were,  they  would 
neither  constitute  a  difficulty  nor  offer  any  basis  for  an  objection  to  that 
schema. 

E.  The  "Industrial  Revolution"  of  the  Twenties. — These  processes 
were  so  entirely  normal  in  the  sense  of  conforming  to  expectation  from 
our  model  and  so  obviously  repeated  the  history  of  preceding  Kondratieff 
downgrades  that  no  war  effects  or  other  disturbances  availed  to  obliter- 
ate the  fact,  and  that  recalling  a  few  familiar  features  suffices  to 
establish  it. 

First,  we  should  not  expect  to  find  fundamentally  new  things,  but  rather 
induced  and  completing  development  on  lines  chalked  out  before  and 
attended  by  strong  increase  in  quantities,  marked  improvement  in  qual- 
ities, "rationalization"  all  around,  an  indefinite  number  of  individually 
small  innovations  producing  a  wide  variety  of  new  specialties,  the  phe- 
nomena which  we  have  called  conquest  of  new  economic  space.  This  is 
what  we  find.  The  electrical,  chemical,  and  automobile  industries,  which 
together  with  their  subsidiaries  and  all  that  directly  and  indirectly  hinges 

1  So  did  public  construction,  of  course.  For  example,  local  authorities  spent,  alone 
through  their  gas,  water,  and  electric  departments,  about  17.5  million  pounds  in  1924. 
Highways,  harbors,  wharves,  docks,  canals,  sewers,  embankments  account  for  another  12.9 
millions;  roads  and  bridges,  for  36.7. 


754  BUSINESS  CYCLES 

upon  them — the  motorcar,  for  instance,  is  responsible  for  a  great  part  of 
the  total  of  postwar  construction:  roads,  garages,  gasoline  stations,  sub- 
urban residences — account  for  90  per  cent  of  the  postwar  changes  in  the 
industrial  organism  and  for  most  of  the  increase  in  real  income.  They 
realized  the  possibilities  created  in  the  Kondratieff  prosperity,  continued 
to  push  ahead  from  the  bases  laid  before,  and  by  so  doing  shaped  things 
into  a  Kondratieff  recession.  So  did  not  only  those  subsidiaries,  such  as 
oil  and  rubber,  but  also  the  minor,  though  still  important,  novelties,  such 
as  steel  alloys,  aluminum,  rayon,  large-scale  retailing,  and  the  organiza- 
tional and  financial  complement — persistence  of  the  merger  movement, 
power  finance  and  so  on.  There  were  exceptions,  as  there  were  in  the 
two  previous  Kondratieff  downgrades,  but  none  of  them  was  quantita- 
tively significant.  The  most  important  one  was  air  transport  on  a  com- 
mercial scale,  which  may  bear  comparison  with  the  role  of  railroads  in  the 
thirties  and  of  electricity  in  the  eighties  of  the  nineteenth  century. 

Second,  we  find  all  the  general  features  which  analysis  and  historical 
observation  have  taught  us  to  associate  with  Kondratieff  downgrades. 
This  will  become  clear  beyond  doubt  in  our  discussion  of  time  series 
which  is  to  follow,  but  it  should  be  clear  independently  of  it,  that  those 
features  can  be  accounted  for  in  terms  of  the  system's  absorption  of  and 
reaction  to  the  new  quantities  and  new  methods.  We  find  prevalence  of 
unemployment1  that  was,  however  many  other  circumstances  may  have 
contributed  to  it,  basically  "technological."  There  was,  though  also 
accentuated  by  other  circumstances,  that  excess  capacity2  which  is 
inseparable  from  the  process  of  rapid  reorganization  of  the  industrial 
apparatus  and  coexists  with  vigorous  expansion  of  output.  We  observe 
that  desperate  struggle  of  firms  for  outlets3  and  against  competition  and 
the  sagging  of  prices  incident  to  the  insertion  of  new  quantities  and  capaci- 
ties, which  understandably  creates  the  picture  of  apparently  permanent 

1  On  this  and  the  two  following  points,  see  infra,  Sec.  F,  I,  a,  2  and  8. 

2  As  has  been  pointed  out  before,  during  a  process  of  expansion  and  rearrangement  of 
this  kind,  normal  utilization  of  capacity — if  it  be  possible  to  speak  of  normals  at  all  in 
situations  which,  though  moving  toward  normality,  are  yet  far  removed  from  any  normality 
— must  always  be  considerably  under  100  per  cent,  even  apart  from  oligopolistic  situations; 
and  this  does  not  necessarily  imply  any  waste,  much  less,  at  any  rate,  than  would  an 
attempt  "to  maintain  the  profitable  operation  of  moribund  plants."     See  M.  C.  Rorty, 
The  Equation  of  Economic  Balance,  Harvard  Business  Review,  April  1934.     To  repeat, 
conditions  in  the  production  of  raw  materials  reflect  the  same  phenomenon  in  the  form  of 
"overproduction.'*     With  varying  degrees  of  severity,  such  conditions  prevailed  through- 
out the  period. 

8  One  symptom  of  this  was  the  orgy  of  advertizing.  Expenditure  on  newspaper  adver- 
tizing alone  has  for  this  country  been  estimated  at  a  billion  and  a  hah*  for  1927,  see  Federal 
Trade  Commission,  Resale  Price  Maintenance  II,  1981.  There  is  no  possibility  of  reliable 
comparison  with  prewar  times,  but  the  fact  of  steady  increase  of  that  item  of  expenditure 
frpm  1922  on  seems  beyond  doubt. 


1919-1929  755 

"overproduction"  or  "overinvestment"  and  the  characteristic  outcry 
about  people's  inadequate  power  or  willingness  to  spend.  And,  masked 
and  retarded  by  resistance  to  adjustments,  the  competing-down  process 
is  clearly  recognizable  both  within  the  relatively  new  and  as  between  new 
and  old  industries,  railroads1  and  coal  being  conspicuous  instances  of  the 
latter.  All  of  which  accounts  for  much  of  the  social  and  business 
atmosphere  of  the  period,  including  its  economic  slogans. 

Even  in  the  comparatively  "pure"  case  of  the  United  States,  "dis- 
turbing factors"  must  be  taken  into  account.  In  the  cases  of  England 
and  Germany  they  are,  of  course,  very  much  more  important.  Where 
they  work  against  our  process  they  are,  qualitatively  at  least,  not  difficult 
to  identify,  and  there  is  nothing  objectionable  in  saying,  for  example, 
that  international  restriction  of  rubber  production  partly  stands  instead 
of  that  fall  in  rubber  prices  which  would  otherwise  have  ensued  and  that 
its  very  existence  proves  the  tendency  which  to  some  extent  it  succeeded 
in  paralyzing.  Where,  however,  external  factors  tend  to  produce  effects 
similar  to  those  that  the  prevailing  phase  of  the  cyclical  process  tends  to 
produce  of  itself,  the  difficulty  sometimes  spreads  from  measurement  to 
qualitative  diagnosis.  We  shall  meet  several  instances  of  this  throughout 
the  rest  of  our  way,  the  most  controversial  of  them  being  the  question 
of  the  effects  of  recovery  policy.  Here  a  single  example  may  suffice. 
We  have  seen  that  in  Germany  factors  extraneous  to  our  process  made 
for  what  was  called  a  consumers'  prosperity.  But  consumers'  pros- 
perities are  also  part  and  parcel  of  Kondratieff  recessions,  and  it  can  be 
shown  that  some  of  the  symptoms  which  that  term  is  intended  to  cover, 
would  have  emerged,  perhaps  not  much  less  strikingly,  in  the  absence  of 
those  factors.  Total  effects  are  difficult  to  allocate. 

1.  English  industry,  as  will  be  evident  from  the  behavior  of  English 
time  series  (see  infra,  Sec.  F,  I)  displays  the  characteristic  features  of 
the  period  much  less  markedly  than  American  or  German  industry. 
This,  as  has  been  explained  in  the  preceding  sections,  must  be  understood 
from  the  sociology  of  her  case,  from  the  way  in  which  England's  economic 
position  was  affected  by  the  war,  and  from  developments  in  other  coun- 
tries that  deprived  her  of  economic  elbowroom.  The  latter  two  factors 
show  in  the  fortunes  of  her  export  trade,  which  recovered  slowly  from  the 
drastic  fall  in  1921 — exports  of  manufactures  and  coal  amounted  to 
1,220  million  pounds  in  1920  and  to  about  half  that  sum  in  1921 — and 
never  reached  the  physical  volume  of  1913  again.  The  latter  increased 
from  1925  to  1929  by  a  little  more  than  8  per  cent  (values  in  current 
pounds  fell),  but  this  did  not  prevent  a  contribution  of  about  25  per  cent 

1  For  England,  cf.  Mr.  Gilbert  Walker's  exceedingly  interesting  study  on  the  Economics 
of  Road  and  Rail  Competition,  Economic  Journal,  June  1988.  A  vast  research  program 
lies  in  this  direction. 


756  BUSINESS  CYCLES 

to  the  unemployment  of  the  latter  year.1  Persistently  favorable  terms  of 
trade — due  to  the  fall  in  the  prices  of  raw  materials,  and  in  keeping  with 
cyclical  expectation — must  no  doubt  be  set  against  this  result,  which 
they,  of  course,  helped  to  produce. 

The  task  indicated  by  this  development  and  by  the  presence  of  large 
areas  that  were  depressed  for  this  and  other  reasons — reorientation  of 
production — was  a  specifically  English  one.  It  provided  the  scope  for 
the  geographical  migration — to  the  metropolitan  area  and  the  South — 
and  for  an  economic  migration  of  resources  into  the  diversified  industries 
catering  for  the  middle  and  lower  strata  of  domestic  consumption,  a 
process  which,  since  as  usual  the  New  did  not  grow  out  of  the  Old, 
accounts  for  bright  and  dark  spots  alike.  Within  this  picture,  however, 
and  barring  the  influence  of  subsidized  building — partly  conditioned  in 
turn  by  the  geographical  migration,  especially  in  the  second  half  of  the 
period — the  role  of  the  outstanding  leaders — electricity,  chemistry,  and 
motorcars — is  obvious.  The  motorcar  industry,  greatly  helped  by  the 
McKenna  duties  and  road  building,  did  much  better  than  in  Germany 
and  substantially  contributed  to  England's  new  industrial  structure  in 
the  West  Midlands  and  elsewhere.  At  first,  prospects  were  most  dis- 
couraging. During  the  war  the  plants  of  motor  (and  cycle)  concerns 
had  been  much  expanded,  but  on  antiquated  lines.  They  emerged  in 
a  state  of  inefficiency  and  immediately  lost  domestic  ground  as  well  as  a 
great  part  of  their  export  to  America.  Moreover,  engineering  firms  whose 
productive  apparatus  was  capable  of  being  adapted  to  the  production  of 
motorcars  then  crowded  into  the  market,  which  was,  finally,  upset  by 
the  sales  of  the  War  Disposals  Department.  All  this  was  an  element  in 
the  crisis  of  1921.  But  a  new  development  started  in  1923  with  the  mass 
production  of  cheap  cars.  From  1921,  when  but  40,000  private  cars 
and  commercial  vehicles  were  produced,  output  rose  steadily,  reaching 
209,000  units  in  1927.2  The  number  of  insured  persons  engaged  in  pro- 
duction and  repair  of  motor  vehicles,  cycles,  and  aircraft  rose  from 
192,000  in  July  1923  to  233,000  in  July  1927,  while  exports  of  complete 
motor  vehicles  and  especially  of  chassis  far  surpassed  the  1913  numbers. 
Some  of  the  subsidiary  industries  and  the  aviation  industry  developed 
still  more  vigorously.  It  is  worth  while  mentioning  that  until  1925-1926 
— when  motor  vehicles  had  increased  by  200  per  cent  over  1919-1920 — 
this  went  on  amidst  what,  to  fit  their  theories,  some  observers  described 
as  "unrelieved  gloom. "  The  nature  of  the  commodity,  more  than  the 
quantitative  importance  of  direct  and  indirect  expenditure  on  it  and 
induced  by  it,  disposes  of  this  diagnosis. 

1  See  Mr.  C.  G.  Clark's  Statistical  Studies  Relating  to  the  Present  Economic  Position 
of  Great  Britain,  Economic  Journal,  September  1931,  for  a  discussion  of  the  various  problems 
concerning  exports. 

8  The  figures  are  as  published  by  the  Society  of  Motor  Manufacturers  and  Traders. 


1919-1929  757 

The  motorcar,  aviation,  and  other  expanding  industries  created  new 
demand  for  steel.  But  the  iron  and  steel  industry  was  nevertheless — 
and  in  spite  of  the  advantages  it  retained:  good  coke,  low  cost  of  trans- 
portation for  some  of  the  world's  highest  grade  ores — ailing  all  the  time 
and  must  be  listed,  along  with  coal  and  cotton,  among  the  weak  spots 
of  the  organism.  Expansion  during  the  war,  depressed  conditions  in 
the  shipbuilding  industry,  loss  of  foreign  markets — directly  and  indirectly 
almost  three-quarters  of  the  United  Kingdom  output  used  to  be  exported 
— account  for  low  prices  and,  in  an  old  industry  which  follows  rather  than 
leads,  for  the  persistence  of  antiquated  methods.  But  it  is  also  true 
that  there  was  little  of  large-scale  domestic  investment  in  industrial  plant 
and  machinery.  Even  the  well-marked  rise  of  steel  consumption  in  1927 
left  it,  though  nearly  30  per  cent  above  the  1913  figure,  distinctly  low.1 

We  pass  by  chemistry — although  it  affords  the  best  example  of  entre- 
preneurial achievement  on  modern  lines  England  has  to  show  in  that 
period — and  everything  else,  in  order  to  note  the  fundamental  role 
electricity — again,  barring  building — played  directly  and  by  what  it 
induced.  Development  in  this  line,  not  a  mere  function  of  growth  but 
originating  and  propelling,  carried  output  of  the  public  supply  system 
of  Great  Britain  from  2.5  billion  kilowatt-hours  to  11  billions  at  the  end 
of  our  period,  when  about  327  million  pounds  had  been  invested  in  the 
generation,  transmission,  and  distribution  of  current.  This  is  less  than 
two-thirds  of  the  cost  per  kilowatt  installed  in  the  United  States.  Tech- 
nical improvement  and  larger  generating  stations  (average  capacity  of 
generating  stations  12,000  kilowatts,  as  against  7,350  in  the  United  States) 
account  for  this  as  much  as  does  the  difference  in  price  level  and  the  lower 
capital  cost  of  steam-power  plant.  Average  consumption  of  coal  per 
unit  generated  fell,  much  as  in  this  country,  from  about  3.2  pounds  in 
1920  to  1.8  in  1930.  English  industry,  which  in  1914  had  still  resisted, 
was  definitively  converted,  and  especially  the  many  small  and  diversified 
industries  that  were  newly  cropping  up  were  electrified  from  the  first. 

But  private  enterprise  supplied  impulse  and  initiative  only  for  the — 
very  considerable — development  of  the  electrotechnical  industries, 
which,  in  the  fashion  of  nineteenth-century  capitalism,  grew  in  a  large 
number  of  concerns  until  the  great  amalgamation  into  the  Associated 
Electrical  Industries  Ltd.  (1929)  put  a  more  modern  touch  upon  it. 
In  the  power  development,  the  initiating  role  belonged  to  the  state,  which 
thus  led  in  the  two  most  important  lines  of  advance  that  characterized 
the  period.  The  significance  of  this  lies  in  the  two  facts,  that  entrepre- 
neurial initiative,  right  from  the  beginning  of  the  Kondratieff,  signally 

1  Home  consumption  of  crude  steel  in  1918  was  2,181,000  tons  quarterly  average;  in 
1927,  2,695,000;  in  1928,  2,417,000;  in  1929,  2,662,000;  a  sharp  fall  thereafter  until  the 
recovery  from  the  second  quarter  of  1983.  Cf.  London  and  Cambridge  Economic  Service. 


758  BUSINESS  CYCLES 

failed  in  an  obvious  and  purely  economic  task,  and  that  public  agencies, 
stepping  into  the  breach,  attacked  it  with  perfect  technological  and 
economic  success,  although,  to  be  sure,  with  the  entrepreneurial  success  in 
America  before  their  eyes.  Only  27  per  cent  of  the  production  of  the  pub- 
lic supply  system  was  privately  owned  at  the  end  of  our  period.  The 
National  Grid,  recommended  in  1925,  was  made  an  item  of  public  policy 
in  1926  (Electricity  Supply  Act). 

Our  sketch  and  the  analysis  underlying  it  can  easily  be  verified  by 
a  list  of  the  industries  that  expanded — or,  expanded  more  than  others1 — 
during  the  decade.  Building,  electricity — with  constructional  and 
electrical  engineering  and  many  subsidiaries  of  both — and  motor  vehicles 
(also  aircraft)  are  there.  Furniture  follows  in  the  wake  of  residential 
construction,  as  does  heating  and  ventilating  apparatus.  Rayon,  of 
course,  was  still  an  innovating  industry.  Presence  of  cement  and  mis- 
cellaneous metals  will  not  surprise  us,  nor  will  the  presence  of  public 
works.  But  then  we  find,  testifying  to  the  reorientation  mentioned  above 
and  characteristic  of  downgrade  developments,  a  great  expansion  in  the 
miscellaneous  trades  and  services,  also  in  professional  services  and  com- 
modity distribution,  in  tobacco,  food,  drink,  silk  •  •  •  while  coal,  other 
engineering,  cotton,  iron  and  steel,  shipbuilding,  and  railways  supply  the 
complement. 

A  postwar  boom  in  business  and  on  the  stock  exchange,  studded  with 
strikes,  started  in  1919  and  lasted  into  the  summer  of  1920.  While  it  was 
going  strong,  banks  hardly  lived  up  to  the  idea  some  economists  have 
formed  about  their  excessive  bent  for  restriction.2  Nor,  as  has  been 
pointed  out  in  a  preceding  section,  was  government  policy  conspicuous  for 
retrenchment.  No  further  comment  is  required  about  the  crash  in  1921, 
which  was  for  England  accentuated  by  its  effects  on  her  foreign  trade  and 
which,  to  repeat,  coincides  with  what  according  to  our  count  would  have 
been  the  beginning  of  a  Juglar  depression.  There  was  some  recovery, 
in  business  as  well  as  on  the  stock  exchange,  in  the  fall,  but  the  ground 
gained  was  lost  in  the  first  half  of  1922.  Then  building  activity  asserted 
itself,  the  Kahn  effect  (secondary  employment)  of  which  presumably 
helped  to  mitigate  things  in  1923.  But  there  were  other  bright  spots, 
the  wool  industry  being  one  of  them,  and  the  stock  exchange  was  active. 
The  year  1924  displayed  all  the  symptoms  of  a  recovery  weighed  down 
by  unfavorable  environmental  conditions.  Subsidized  building  lent 

1The  criterion  is  employment — no  doubt  an  unreliable  guide,  the  short-comings  of 
which  account  for  the  absence,  excepting  rayon,  of  the  chemical  industry,  but  which  may 
pass  muster  for  the  purpose  in  hand. 

2  The  severity  of  the  subsequent  breakdown  illustrates  well  the  consequences  of  "liberal" 
credit  policies.  It  should  be  particularly  noticed  that  the  Bank  of  England  gave  an  impulse 
in  this  direction  by  repaying  Special  Deposits.  See  S.  E.  Harris,  Monetary  Problems  of  the 
British  Empire,  1931,  Chap.  III. 


1919-1929  759 

support,  however.  Imports  increased,  prices  rose  somewhat,  unemploy- 
ment decreased.  Even  the  shipbuilding  and  cotton  situations  were 
relieved.  1925  brought  expansion  in  many  of  the  newer  industries,  but 
what  might  have  been  the  completion  of  recovery  was  greatly  interfered 
with  by  labor  difficulties  (coal,  shipping,  wool)  and  the  numerous  partic- 
ular depressions  in  individual  industries,  i.e.,  depressions  due  to  causes 
peculiar  to  them. 

The  clearly  abnormal  situation  of  1926  makes  the  year  difficult  to 
class.  A  struggle  that  in  any  other  country  might  have  spelled  revolution 
paralyzed  everything.  The  difficulty  extends  to  1927,  because  the 
repairing  of  the  omissions  of  1926  and  the  fact  that  1927  marks  the  peak 
of  subsidized  building  would  in  themselves  suffice  to  explain  the  brighter 
colors  of  that  year.  However,  we  do  not  rely  on  aggregates  and  indices 
only,  but  also  and  primarily  on  what  happened  in  the  industrial  organism. 
There  many  of  the  new  things  mentioned  above  now  began  to  gather 
momentum.  It  does  not  seem  unreasonable,  therefore,  to  speak  of  a 
new — the  fourth — Juglar,  since  this  is  not  more  than  our  way  of  expressing 
those  very  facts.  And  it  will,  perhaps,  also  seem  acceptable  to  date  it 
for  the  purpose  of  counting  from  1926,  seeing  that  in  the  absence  of  a 
clearly  abnormal  event  the  prosperity  phase  would  presumably,  condi- 
tions in  1925  having  been  what  they  were,  have  set  in  earlier.  There  is 
no  reason,  however,  why  we  should  insist  on  this. 

2.  In  Germany,  the  term  Rationalization  was  used  more  commonly 
than  it  was  anywhere  else,  in  order  to  describe  the  industrial  processes  of 
the  postinflation  period.  It  not  only  expressed  what  amounted  to  a 
conscious  national  effort  sponsored  by  all  classes — though  partly  from 
different  motives — and  encouraged  by  the  federal  government,1  but  it  also 
expresses  the  gist  of  what  we  mean  by  downgrade  developments :  exploita- 
tion to  the  utmost,  partly  under  duress,  of  existing  possibilities  of  tech- 
nological and  organizational  innovations  on  lines  and  principles  estab- 
lished before  but  steadily  improved  in  the  process;  revision  of  the  whole 
structure  of  industry  in  quest  of  increased  efficiency;  systematic  struggle 
with  each  item  of  the  list  of  costs — all  of  which  is  exemplified  to  perfection 
by  the  postwar  history  of  all  branches  of  German  industry.  Few  things 

1  M ore  burocratico,  this  encouragement  found  expression  in  a  new  governmental  board 
called  the  Reichskuratorium  fair  Wirtschaftlichkeit,  the  first  achievement  of  which  con- 
sisted, characteristically  enough,  in  defining  the  concept  of  rationalization.  It  is,  however, 
more  important  to  note  that  the  spokesmen  of  the  trade  unions  were  by  no  means  hostile. 
They  sometimes  maneuvered  themselves  into  somewhat  difficult  positions  by  arguing  that 
high  wages  would  make  themselves  possible  by  enforcing  rationalization  and  otherwise 
espousing  the  cause  of  technological  improvement,  while,  of  course,  allegiance  was  at  the 
same  time  due  to  the  doctrine  that  such  improvement  necessarily  injures  the  interests  of 
labor.  Capitalism  was  condemned  for  innovating  too  slowly  and  too  quickly  at  the  same 
time.  But  no  actual  resistance  was  offered  on  the  latter  ground. 


760  BUSINESS  CYCLES 

were  fundamentally  new,  the  most  important  items  being1  the  production 
on  a  large  scale  of  synthetic  nitrogen  (the  Leunawerke  of  the  Dye  Trust 
were,  however,  erected  in  1916  as  a  war  measure;  the  Haber-Bosch  inven- 
tion dates  from  1913),  the  production  of  aluminum,  which  in  Germany 
also  dated  from  the  war,  and,  on  a  commercial  scale  at  least,  the  radio 
and  the  aeroplane.  Nevertheless,  there  was  as  complete  a  transforma- 
tion of  the  economic  organism  as  there  was  in  England  after  the  Napo- 
leonic wars. 

For  purposes  other  than  ours  it  would  be  necessary  to  dwell  on  the 
great  and  increasing  role  played  by  the  industrial  activity  of  the  Reich, 
the  states,  and  the  municipalities.  As  far  as  the  Reich  was  concerned,  it 
had  acquired  most  of  its  business  property  during  the  war  and  with  no 
intention  beyond  providing  for  war  requirements,  although  both  the  logic 
of  the  situation  thus  created  and  the  spirit  of  the  times  made  it  easier  to 
expand  than  to  liquidate  it.  But  the  states  had  inherited  important 
interests  from  their  monarchic  predecessors  and  added  to  them  as  a  matter 
of  principle,  while  the  municipalities  simply  went  further  along  the  path 
of  "  municipal  socialism."  For  1925  total  turnover  of  all  public  enterprise 
was  estimated  by  Dr.  Marschak  at  10  billion  marks2  and  total  value  of 
all  publicly  owned  property  by  Professor  J.  Hirsch — conservatively — 
at  52  billions,  roughly  one-fifth  of  the  national  total.  However,  26  bil- 
lions are  accounted  for  by  the  federal  railways — which  were  merely 
transferred,  as  a  consequence  of  postwar  arrangements,  from  the  states 
that  previously  owned  them  to  a  corporation  owned  by  the  Reich — and 
1  billion  by  the  federal  post  office.  The  other  concerns  mainly  or  wholly 
owned  by  the  Reich  were  combined  into  a  holding  company  (Vereinigte 
Industrie-Unternehmungen  A.G.)  and  included  electrical  and  electro- 
technical,  aluminum — the  bulk  of  the  German  aluminum  production  was 
controlled  by  the  Vereinigte  Aluminium  Werke  wholly  owned  by  the 
Reich — nitrogen,  iron-  and  steelworks,  and  mining.  The  Reich  also 
founded  a  commercial  bank  of  first  rank,  the  Reichskredit-Gesellschaft.3 

Similarly,  the  states  strengthened  their  foothold,  especially  in  the 
fields  of  electricity  and  mining,  but  also  in  others.  The  municipalities 
built  and  operated  additional  gas,  water,  electrical  works,  street  railways, 
slaughterhouses,  and  so  on,  but  did  not  materially  go  beyond  utilities  and 
residential  building.  The  interesting  thing  about  those  industrial 

1  We  do  not  list  the  hydrogenation  of  coal  (or  the  staple  fiber)  because  it  did  not,  during 
the  twenties,  play  any  significant  role,  though  the  technological  bases  were  being  laid  for 
the  developments  of  the  thirties.     This  was  done  within  the  Dye  Trust  (Bergius  process). 

2  See  Wirtschaftsdemokratie,  ed.  by  the  Allgemeine  Deutsche  Gewerkschaftsbund,  3d 
ed.,  1929,  Appendix  I. 

3  The  states  and  other  public  bodies  did  not  lag  behind.     It  is  interesting  to  note  that 
total  assets  of  banks  controlled  by  public  bodies  were,  in  1927, 12.7  billion  marks  as  against 
the  16.1  billions  of  all  other  banks. 


1919-1929  761 

properties  of  the  Reich  and  the  states  is  the  thoroughly  businesslike 
manner  in  which  they  were  managed.  The  public  authority  was  a  stock- 
holder, sometimes  but  not  necessarily  the  only  or  the  controlling  one,  and 
interfered  but  little.  The  managements  enjoyed  not  much  less  inde- 
pendence than  they  would  have  in  any  ordinary  case,  and  they  behaved 
as  managements  ordinarily  do — they  took  pride  in  technological  perfec- 
tion, good  profits,  and  sizable  reserves.  The  whole  arrangement  seemed 
devised  in  order  to  put  competent  businessmen  in  charge  and  to  keep 
politics  at  bay.  As  a  possible  solution  of  the  problem  of  industrial 
leadership  in  a  socializing  state — a  problem  that  loomed  large  in  the 
discussions  of  the  Commission  on  Socialization  in  1919 — it  seems  to  merit 
attention. 

For  our  purpose,  however,  we  need  not  take  account  of  this  element  in 
the  economic  life  of  Germany,  except  in  the  case  of  the  production  of 
electrical  power  which,  vigorously  developing  even  during  the  period  of 
inflation,  reached  20.3  billion  kilowatt-hours  in  1925,  25.1  in  1927,  and 
30.7  in  1929.  Of  this  total,  from  2  to  4  billions  were  produced  in  hydro- 
electric plants.  Lignite1  was  used  in  the  production  of  about  one-third 
of  it.  About  70  per  cent  of  the  output  of  public  supply  stations — not  of 
the  totals  mentioned — was  controlled  by  either  the  Reich  or  the  states  or 
by  municipalities,  although  the  relative  share  of  the  latter  fell,  as  com- 
pared with  prewar  times,  owing  to  the  improvements  in  long-distance 
transmission,  which  was  unfavorable  to  local  sources  of  moderate  size. 
The  federal  governments  power  concerns  and  participations  in  power 
concerns,  of  which  the  A.G.  fiir  Elektrizitatswirtschaft  and  the  Rheinisch- 
Westfaelisches  Elektrizitatswerk  were  the  most  important,  were  paralleled 
by  the  Preussische  Elektrizitats  A.G.,  which  combined  the  power  inter- 
ests of  the  Prussian  state,  and  by  the  works  owned  by  Saxony,  Bavaria, 
Baden,  and  Thuringia.  This  alone,  and  apart  from  the  municipal  sta- 
tions, practically  amounted  to  public  control  of  generation  and  transmis- 
sion and,  of  course,  foreshadowed  progress  toward  a  perfectly  coordinated 
system  of  interconnecting  superpower  stations.  The  work  of  electrifica- 
tion was,  however,  not  nearly  completed  even  in  industry,  while  agricul- 
ture in  1925  consumed  only  0.4  billion  kilowatt-hours  of  electrical  energy, 
and  railways  were,  even  in  1928,  electrified  only  to  2.4  per  cent  of  track- 
age. The  electrification  of  the  household  was  hardly  begun.  As  far  as 
the  writer  is  able  to  make  out,  total  investment  was,  1925  to  1929,  of  the 
order  of  magnitude  of  2  billion  marks.2 

1  Brickett  production  for  domestic  fueling,  the  use  of  lignite  as  a  chemical  raw  mate- 
rial, and  its  use  for  the  production  of  electric  current  account  for  its  spectacular  career. 
Mechanization  of  lignite  mining,  of  course,  helped.     Horsepower  installed  rose  from  55,000 
in  1895  to  209,000  in  1907  and  766,000  in  1925;  output,  which  increased  even  during  the 
war,  from  87.2  million  metric  tons  in  1913  to  139.7  in  1925  and  165.6  in  1928. 

2  Assuming  that  investment  per  kilowatt  installed  was  about  1,000  marks. 


762  BUSINESS  CYCLES 

Value  of  output  of  the  electrotechnical  industry  made  a  big  stride 
between  1925,  when  it  was  2.1,  and  1927,  when  it  was  2.7  billion  marks.1 
These  are  very  hazardous  estimates,  but  they  are  fully  borne  out  by  the 
official  and  more  reliable  figures  of  value  of  exports.  They  rose  only  by 
little  more  than  20  per  cent  during  those  years3 — to  441.2  millions  or 
26.7  per  cent  of  the  total  for  all  countries,  which  compares  with  the 
417.5  million  marks  and  a  25.2  per  cent  share  of  the  United  States.  But 
if  that  was  possible  in  the  face  of  the  existing  barriers  and  the  fact  that 
export  in  this  line  requires  a  great  deal  of  capital,  it  is  safe  to  assume  that 
domestic  sales  increased  much  more  strongly.  Innovation  was  effec- 
tively corraled,  however.  There  were  new  concerns,  but  the  two  that 
had  been  leading  before  the  war,  Siemens  and  the  General  Electric, 
retained  that  position. 

In  1913  the  machinery  industry — excluding  electrotechnical  produc- 
tion and  boilers  but  including  locomotives — worked  at  practically  100 
per  cent  capacity,  producing  products  of  a  value  of  2.7  billion  marks 
(postwar  territory)  to  which  export  contributed  0.74  billions.  In  1925 
value  of  production  at  prewar  prices  has  been  estimated  at  about  1.9  bil- 
lions while  capacity  was  for  3.36. 3  Progress  during  1925-1929  in  quantity 
and  quality  of  output  was  very  great,  value  of  products  at  current  prices 
rising  by  38  per  cent,  considerably  more  than  anywhere  else,  and  exports, 
which  rose  to  about  1  billion  marks  by  1928,  recovering  part  of  their  pre- 
war position.4  While  in  many  respects  this  was  a  great  entrepreneurial 
achievement — or  rather,  the  result  of  a  long  series  of  entrepreneurial 
achievements  within  a  large  number  of  medium-sized  firms — which  con- 
tributed substantially  to  the  economic  processes  of  the  time,  and  while, 
though  of  smaller  quantitative  importance,  the  industries  producing 
optical,  medical,  and  other  instruments,  photographic  apparatus,  and 
so  on — typewriters  have  been  included  in  machinery — kept  well  in  step, 
the  automobile  industry  stagnated,  notwithstanding  prohibition  of 

1  Comparison  would  be  greatly  to  the  disadvantage  of  the  United  States,  where  that 
value  rose  but  moderately  during  the  same  time.     This  may,  however,  be  as  much  due  to 
the  faultiness  of  estimates  as  to  the  fact  that  in  this  country  the  big  stride  had  been  made 
before  1925. 

2  However,  1928  brought  a  great  increase,  although  Germany's  share  in  the  world's 
trade  still  remained  far  below  the  50  per  cent  of  1913. 

3  Cf.  the  Denkschrift  tiber  die  Maschinenindustrie  der  Welt,  October  1926,  prepared 
for  the  League's  International  Economic  Conference  by  the  Association  of  German  Pro- 
ducers of  Machines  (Verein  Deutscher  Maschinenbau-Anstalten).     With  due  respect  for 
this  excellent  memorandum,  to  which  the  writer  is  much  indebted  for  a  better  grasp  of  the 
situation  of  that  industry  in  1925,  exception  must  be  taken  to  the  two  methods  suggested, 
p.  17,  for  arriving  at  an  estimate  of  capacity.     For  the  estimates  actually  made  it  may  be 
urged,  however,  that  the  officers  of  the  Association  were  likely  to  have  a  pretty  good 
impression. 

4  United  States  exports  were  1,688  million  marks  in  1928. 


1919-1929  763 

imports  followed  by  heavy  protection,  until  about  1927,  not  only  because 
of  obvious  postwar  difficulties  but  also  because  of  a  temporary  inability 
to  modernize  itself  which  is  not  completely  explained  by  those  difficulties. 
The  old  firms,  which  before  the  war  had  been  among  the  international 
leaders,  lingered  on  or  died,  and  only  inefficient  new  ones  were  added  for  a 
time  until  the  belated  advent  of  the  small  cheap  car  that  was  sufficiently 
economical  as  regards  taxes  and  gasoline.  Domestic  progress  was  also 
quickened  by  General  Motors  acquiring  and  Ford  erecting  plants  in 
Germany,  prices  eventually  fell  to  a  level  on  which  the  homemade  car 
had  a  chance  to  sell — 1928,  the  year  of  maximum  importation,  is  the 
first  year  in  which  this  can  be  said  to  have  been  the  case — and  the  weak- 
est firms  dropped  out.  We  therefore  note  that  both  investment  and  sales 
began  to  be  significant  in  the  second  half  of  our  period;  but,  stated  in 
absolute  figures,  success  remained  small.  To  the  20,000  units  which, 
including  trucks  and  buses,  were  produced  in  1913  corresponded  not 
quite  63,000  in  1925  and  about  138,000  in  1928.  Motor  cycles  did  better. 

Mechanization  (e.g.,  of  rolling),  concentration  of  production  in  bigger 
units  optimally  located  (e.g.,  Hamborn),  varied  progress  in  electrometal- 
lurgy, standardization  of  products,  and  improvements  in  the  use  of  heat 
and  power  are  the  familiar  features  of  typically  induced  developments  in 
mining  and  the  heavy  industries.  We  will  confine  ourselves  to  the 
organizational  aspect.  It  has  been  pointed  out  in  Chap.  VII  that  infla- 
tion gave  a  powerful  impetus  to  the  merger  movement,  which  already 
had  gone  very  far  during  the  Kondratieff  prosperity.  The  losses  of  terri- 
tory, often  cutting  through  the  domains  of  concerns  and  upsetting 
established  relations  between  materials  or  stages  of  production,  also 
necessitated  reorientation.  Many  unshapely  and  unmanageable  mon- 
sters resulted,  which  were  unable  to  live  as  soon  as  the  contours  of  reality 
emerged  from  the  fog.  The  breakdown  of  the  Stinnes  concern,  the  dis- 
solution of  the  Siemens-Rheinelbe-Union  are  but  conspicuous  instances 
of  a  process  of  liquidation  which  was  referred  to  as  the  "crisis  of  concerns"1 
and  which,  even  where  it  did  not  lead  to  failure,  put  an  end  to  many  vertical 
and  horizontal  combinations,  participations,  understandings,  and  so  on. 
Even  such  concerns  as  Krupp,  Stumm,  and  Rombach,  had  to  retrace 
steps  and  to  take  losses.  The  ordinary  and  the  international  cartel  and 
regulation  by  public  authority  (Zwangssyndizierung)  were  the  remedies 
resorted  to.  As  an  example  for  the  latter  kind  of  rationalization  we  will 
mention  the  potash  industry.2 

In  this  case  it  had  already  started  during  the  war,  when  the  emergence 
of  new  works  and  the  sinking  of  new  pits  was  barred  by  government 

1  The  German  word  Konzern  means  combination  rather  than  firm. 

2  Of.  the  report  of  the  Survey  on  Economic  Conditions  (Wirtschaftsenquete)  on  that 
industry.    This  report  also  appeared  as  a  book,  1929, 


764  BUSINESS  CYCLES 

decree  (1916).  Three  further  decrees  (1919,  1920,  and  1921)  put  a  heavy 
premium  on  the  closing  of  mines,  which  was  also  facilitated  by  the  merger 
movement  in  progress.  While  the  number  of  works  enjoying  quota 
increased  from  1921  to  1928,  when  there  were  229  of  them,  the  number  of 
works  in  operation  in  the  same  period  decreased  to  60  which,  greatly 
improving  methods  and  the  quality  of  the  product — also  82  per  cent  of 
the  product  was  by  1928  no  longer  sold  in  the  crude  state  but  processed 
within  the  industry,  by-products  gaining  steadily  in  importance — and 
reducing  cost,  increased  their  output  more  than  threefold  in  the  course 
of  five  years.1  The  mergers  left  six  independent  concerns  in  business, 
of  which  three  produced  80  per  cent  of  the  national  total.  Sales  were 
centralized  in  the  potash  syndicate,  which  in  1925  negotiated  an  English 
loan  and,  according  to  a  complicated  schedule,  granted  rebates  to  the 
various  (five)  classes  of  domestic  buyers  but  did  not  fix  the  prices.  These 
were  fixed  by  a  public  authority,  the  federal  potash  council  (Reichs- 
kalirat),  and  subject  to  the  approval  of  the  Ministry  of  Economic 
Affairs — needless  to  ask  whether  or  why  they  were  rigid.  In  inter- 
preting this  case,  which  is  of  interest  far  beyond  its  relevance  to  our 
subject,2  the  reader  should,  however,  bear  in  mind  the  fact  that  capital 
invested  had,  before  the  war,  been  estimated — very  unreliably  to  be  sure 
— at  1.4  billion  marks  of  prewar  purchasing  power,  while  in  the  Survey 
of  Economic  Conditions  it  was  for  the  end  of  our  period  estimated  at  from 
600  to  700  million  marks  of  postwar  purchasing  power — a  loss  which  no 
doubt  comes  under  the  headings  of  malinvestment  and  waste  of 
competition. 

Several  other  cases  of  more  or  less  the  same  type — coal3  and  brown- 
coal  mining  for  instance — could  be  cited.  We  will,  however,  confine 
ourselves  to  the  case  of  iron  and  steel.  It  could  also  serve  as  an  example 
of  international  regulation  which  suggested  itself  owing  to  the  great 
expansion  in  capacity  coupled  with  both  improvement  and  more  eco- 
nomical use  of  the  product  all  over  the  world.  Mere  quantities,  very 
clearly  indicating  the  upswing  that  lasted  from  1925  to  1928,  increased 
during  those  years  by  15  per  cent  (iron)  and  20  per  cent  (steel),  in  Europe 
alone  by  respectively  25  and  30  per  cent.  German  production  of  steel 
surpassed  its  prewar  level  of  not  quite  12  million  tons  (1913,  postwar 

1  The  output  of  the  industry  rose  to  about  10  per  cent  above  prewar  level  in  1925  and 
further  increased  in  1928.     Costs  of  power  and  heat  in  terms  of  lignite  fell  to  less  than  half. 
Workmen  employed  fell,  1923  to  1928,  from  40,000  to  19,000.     Wage  rates  rose  over  60  per 
cent  from  1924  to  1928,  so  that  wage  bill  and  labor  cost  per  unit  of  product  fell. 

2  The  measures  taken  and  the  way  in  which  they  worked  out  should  in  particular  be 
interesting  to  American  students  of  planning  of  the  NRA  type.     The  problem  dealt  with 
by  those  measures  was  exactly  analogous  to  the  one  facing  American  policy  in  the  time  of 
the  NRA. 

3  Coal  Act,  setting  up  the  Federal  Coal  Council  (Reichskohlenrat)  of  Apr.  24,  1919. 


1919-1929  765 

territory)  in  1925,  and  was  over  16  million  tons  in  1927,1  though  the 
production  of  foundries  in  1927  was  only  at  the  1913  figure  (exactly). 
The  production  of  rolling  mills  (semifinished  and  finished  products) 
was  in  the  same  year  far  above  the  1913  figure  (by  about  60  and  30  per 
cent  respectively2).  The  Berlin  Institute's  index  of  iron  prices  which  in 
1925  stood  at  125  per  cent  of  1913,  fell  in  1926  to  112  per  cent  and  then 
rose,  with  a  setback  in  1928,  through  1930. 

For  us  the  interesting  point  is  that  that  development  occurred  in  the 
course  of  an  "organizational  rationalization"  which  was  effected  exclu- 
sively by  entrepreneurial  effort  from  within  the  industry  and  is  suggestive 
of  earlier  American  examples.  The  new  Vereinigte  Stahlwerke,  which 
attained  corporate  existence  in  1926,  were  a  unit  of  control  that  aimed  at 
concentration  and  specialization  of  production  in  optimally  located  plant. 
From  1926  to  1933  (when  in  the  wake  of  the  crisis  another  reorientation 
and  reorganization  took  place)  pits  were  reduced  from  48  to  25,  iron- 
works from  140  to  66,  steel  foundries  from  20  to  8,  rolling  mills  from 
17  to  10.  Very  considerable  investment  was  required  to  achieve  this,  and 
almost  immediately  after  their  foundation  the  Stahlwerke  incurred  a 
debt  of  over  500  million  marks,  a  little  more  than  half  of  which  was  spent 
on  the  erection  of  new  plant  and  the  improvement  of  existing  plant. 
Their  quota  in  the  syndicate  was  less  than  40  per  cent,  however.  Not 
only  such  firms  as  Krupp,  Mannesmann,  and  Hoesch  retained  their 
independence,  forming  other  alliances  and  expanding  on  their  own,  but 
also  another  group  was  formed  under  the  leadership  of  the  Rheinelbe- 
Union,  which  embarked  on  an  extensive  investment  program  of  similar 
type3  and  soon  floated  an  issue  of  800  million  marks.  The  works — those 
belonging  to  different  combinations  as  well  as  independent  ones — which 
produced  steel  specialties,  in  1927  formed  an  organization  called  Deutsche 
Edelstahlwerke.  Every  one  of  these  steps  was  accompanied  by  induced 
innovation  of  the  technological  type,  and  all  of  them  created  an  almost 
completely  new  industrial  organism.  That  in  the  subsequent  crisis 
things  should  have  presented  a  picture  which  it  was  as  easy  as  it  was 
superficial  to  describe  in  terms  of  excess  capacity  and  malinvestment  will 
not  surprise  us. 

The  quantitatively  most  important  event  in  the  chemical  industry, 
which  carried  everything  before  it  and  solved  the  problem  created  by 
the  loss  of  patents  after  the  war,  has  been  mentioned  already.  Mark 
volume  of  chemical  production  rose  by  33  per  cent  from  1924  to  1928, 

1  At  the  same  time  the  number  of  workmen  employed  by  steelworks  was  lower  by  about 
16  per  cent. 

2  It  is  for  our  purposes  important  to  note  that  there  was  a  sharp  setback  in  1928. 

8  The  above  refers  to  the  western  district.  The  heavy  industry  of  Upjter  Silesia  com- 
bined independently,  and  the  steelworks  of  central  Germany  (Mitteldeutsche  Stahlwerke) 
formed  yet  another  group  in  which  the  Vereinigte  Stahlwerke  participated  to  50  per  cent. 


766  BUSINESS  CYCLES 

exports  went  over  1  billion  by  1926 — first  indications  of  the  unfathomable 
possibilities  of  synthetic  materials.  No  other  industry  displays  so 
clearly  an  ineluctable  necessity  of  largest  scale  enterprise — the  J.  G. 
Farben  (Dye  Trust)  produces  about  100  per  cent  of  total  German  output 
of  dyes,  about  85  per  cent  of  the  output  of  synthetic  nitrogen,  and  about 
90  per  cent  of  the  output  of  sulphuric  acid,  and  the  Imperial  Chemical 
Industries  Ltd.  in  England,  the  fitablissements  Kuhlmann  in  France, 
Montecatini  in  Italy  and,  to  a  lesser  extent,  Du  Pont  de  Nemours  in  this 
country  hold  comparable  positions — as  well  as  the  necessity,  for  profits 
and  even  for  survival,  of  incessant  innovation,  as  does  this  premier  indus- 
try of  the  future.  Into  its  problems  we  cannot  enter  any  more  than  into 
the  problems  of  the  rayon  industry,  which  during  our  period  made  its 
decisive  stride,  consumption  about  doubling  from  1925  to  1927.  This 
compares  with  all  but  stagnation  in  the  cotton,  wool,  linen,  and  jute 
industries.  The  marked  upswing  in  1927,  which  was  well  sustained 
afterward,  should  be  mentioned,  however.  And  so  should  the  concen- 
tration which  went  on  and  reduced,  for  example,  in  the  cotton  industry 
the  21,600  units  that  existed  in  1907  to  8,000  by  1925.  The  fact  that  the 
limits  set  to  these  fragments  of  a  sketch  do  not  allow  us  to  enter  into  the 
history  of  other  industries  is  particularly  regrettable,  because  it  makes  it 
impossible  to  describe  those  items  which  would  reflect  the  consumers' 
prosperity  of  that  time  and  that  large  medley  of  individually  small 
innovations — many  of  them  little  more  than  successful  insertions  of  some 
new  article  or  brand  by  means  of  dashing  advertizing  campaigns — which 
in  Germany,  as  elsewhere,  contributed  so  much  that  is  most  character- 
istic of  its  processes. 

Dr.  Clausing,1  applying  Professor  Spiethoff's  model,  counts  two 
cycles,  the  first  from  a  trough  in  November  1923  to  another  trough  in 
January  1926,  the  second  from  an  incipient  recovery  in  February  1926 
to  a  depression  (Niedergang)  from  the  second  half  of  1929  to  1932.  From 
our  standpoint  there  is  very  little  to  add  to  or  to  criticize  in  this.  There 
is  no  doubt  that  1927  was  a  year  of  prosperity  also  in  our  sense.  As  the 
reader  will  easily  verify  by  reference  to  what  has  been  said  above,  this 
prosperity  links  up  with  industrial  innovation  which  displays  all  the 
characteristics  we  associate  with  Kondratieff  downgrades.  We  shall  see 
later  that  time-series  evidence  would  support  that:  unemployment 
decreased  strongly;  price  level  rose;  liquidations,  receiverships,  and 

1  Gustav  Clausing,  Die  Wirtschaftlichen  Wechsellagen  von  1919  bis  1932,  1983.  C.  T. 
Schmidt,  German  Business  Cycles.  1924-1933,  published  by  the  National  Bureau  of  Eco- 
nomic Research,  1934,  also  observed  two  cycles,  the  one  from  a  trough  in  December  1923 
to  a  trough  in  March  1926,  culminating  about  March  1925,  the  other  from  approximately 
April  1926  to  the  late  summer  of  1932,  the  crest  being  difficult  to  establish  (pp.  169  and 
170).  He  notes  that  after  the  beginning  of  1927  a  "growing  international  similarity  is 
present."  (Page  248.) 


1919-1929  767 

bankruptcies  fell  to  their  postwar  minima;  foundations  of  new  firms 
increased  and  so  did  corporate  earnings.  It  was  clearly  the  familiar  pic- 
ture of  what  we  are  in  the  habit  of  calling  a  Juglar  prosperity.  Although 
we  have  decided  (Chap.  VII)  not  to  go  on  counting  Juglars  in  the  case 
of  Germany,  it  is  worth  while  to  note  that  it  would  not  be  impossible  to 
do  so.  Unlike  this  country,  and  much  more  than  England,  Germany 
experienced  in  1928  what  most  people  will  agree  to  call  recession,  but  the 
symptoms  do  not  quite  answer  to  our  conception  of  that  phase  and  are 
in  some  respects  contradictory.  The  Berlin  Institute's  index  of  produc- 
tion and  employment  fell,  while  other  factors,  e.g.,  money  in  circulation, 
cost  of  living,  and  wholesale  prices  rose  a  little.  There  was,  however,  a 
rally  at  the  end  of  the  year  which  lasted  through  the  first  four  months  of 
1929,  when  the  system,  slowly  at  first,  began  to  slide  off  into  depression. 
And  the  case  for  expecting  reaction  to  the  preceding  industrial  revolution 
remains  strong. 

That  prosperity  did  not  start  with  1927,  but  was  in  full  swing  at  the 
beginning  of  that  year.  In  terms  of  industrial  innovation  and  investment 
as  well  as  of  many  aggregative  indices,  such  as  stock  prices,  imports  of 
raw  materials,  orders  received,  total  output,  pig-iron  production,  employ- 
ment, the  second  half  of  1926  was  a  time  of  prosperity  in  our  sense  and 
must  be  considered  as  the  beginning  of  the  upswing  discussed  in  the 
preceding  paragraph.  What  went  before,  however,  should  not  be 
interpreted  as  a  cycle  in  the  sense  of  either  our  own  or  any  other  analytical 
schema.  The  year  1924  starts — from  a  crisis  and  heavy  unemployment 
in  the  winter  1923-1924 — with  an  upswing  which  means  little  more  than 
that  business,  having  fallen  from  the  clouds  of  inflation,  was  trying  to 
find  its  feet.  It  was  cut  short  in  the  middle  of  that  year  by  the  effects 
of  taxation  and  of  a  pull  of  the  Reichsbank's  curb.  Prospects  of  foreign 
credits — and  the  first  foreign  credits  actually  given — together  with  sham 
profits  due  to  low  evaluation  of  assets  in  terms  of  the  new  mark  then 
account  for  a  spurt  toward  normal  volumes,  which  lasted  from  about 
July  1924  to  about  February  1925.  A  severe  relapse — in  the  fourth 
quarter,  particularly  severe  in  employment — followed  upon  this,  which 
was  due  to  a  variety  of  circumstances  peculiar  to  the  situation.  Hence, 
it  would  not  do  to  identify  that  up  and  down  as  a  Kitchin. 

3.  In  the  United  States  conformity  to  expectation  during  that  period 
is,  as  stated  above,  so  obvious  as  to  make  it  almost  superfluous  to  prove 
it,  a  fact  the  value  of  which  is  enhanced  by  the — relatively  speaking — 
small  importance  of  external  disturbances  in  our  sense.  That  the  events 
in  the  fields  of  electricity,  motorcars,  and  chemistry  do  not,  in  our  ter- 
minology, constitute  fundamentally  new  but  induced  and  completing 
developments,  which  proceeded  from  bases  laid  in,  roughly,  the  two 
prewar  decades,  needs  additional  emphasis  as  little  as  does  the  fact  that 


768  BUSINESS  CYCLES 

it  was  those  developments  that  "carried"  the  economic  processes  of  the 
period.  We  may,  however,  note  the  substantive  novelty  of  aviation  as  a 
commercial  success — 1925  may  serve  as  a  date — which  was  perhaps  the 
most  important  exception.  The  interesting  thing  about  this  industry  is 
that  it  developed  on  its  own  and  not,  as  might  have  been  expected  from 
standpoints  other  than  ours,  as  an  appendage1  to  an  older,  say  the  auto- 
mobile, industry  in  spite  of  the  similarity  its  problems  bear  to  those  of 
the  latter.  Exactly  as  the  telephone  industry  was  not  built  up  by  the 
telegraph  industry  and  has  shown  no  tendency  to  be  dominated  by  it, 
and  as  the  rise  of  the  automobile  industry  owed  but  little  to  the  carriage 
and  bicycle  industries — or,  we  may  add,  to  the  firms  that  previously 
produced  the  Otto  motors — or  as  the  moving  picture  industry,  which  we 
might  also  list  among  the  genuine  innovations  of  the  period,  did  its  own 
pioneering  and  was  not  the  work,  technically,  financially,  or  commercially, 
of  the  theater  interests,  so  aviation  supplies  another  instance  in  verifica- 
tion of  the  hypothesis  of  New  Firms  and  New  Men  (Chap.  Ill)  arising 
independently  of  the  Old  Firms  and  laying  themselves  alongside  of  them.2 
The  same  often  holds  true  of  new  specialties  within  each  great  line  of 
advance,  as  within  the  field  of  electrical  industries,  partly  at  least,  in 
the  cases  of  the  radio  and  of  the  refrigerator. 

a.  Power  production  increased  from  38.9  billion  kilowatt-hours  in 
1919  to  over  97  in  1929,3  only  1921  marking  a  relapse  of  about  8  per 
cent.  Roughly  95  per  cent  of  this  was  produced  by  privately  financed 
enterprise  and  over  half  of  it  by  the  General  Electric,  Insull,  Morgan, 
Mellon,  Byllesby  and  Doherty  groups  and  a  dozen  corporations  jointly 
controlled  by  these.  Although  the  more  remote  effects  of  this  develop- 
ment on  industrial  activity  in  general  were  much  more  important  factors 

1  This  has  been  stressed  so  well  by  Professor  M.  W.  Watkins,  The  Aviation  Industry, 
Journal  of  Political  Economy,  February  1931,  that  we  cannot  do  better  than  quote  from 
that  paper.     The  rest  of  the  above  paragraph  is  almost  bodily  taken  from  pp.  67  and  68. 

2  Professor  Watkins'  comments  on  the  phenomenon  should  be  quoted:  "The  explana- 
tion seems  to  be  that  the  managers  and  directors  of  older  industries,  once  they  have  suc- 
ceeded in  establishing  as  an  economic  'going  concern'  the  special  branch  of  industry  with 
which  they  are  primarily  identified,  lose  their  adventurous  inclinations.     They  tend  to 
become  skeptical  of  new  processes  and  new  products.     They  lose  the  Vision '  of  industrial 
pioneers.     They  become  absorbed  in  the  complicated  routine  of  their  own  affairs  and  the 
ever  recurring  problems  of  adjustment  and  adaptation  of  which  no  field  of  business  enter- 
prise is  free.     In  these  circumstances,  it  is  only  the  far-sighted,  uneasy ',  venturesome  indi- 
viduals here  and  there  who  are  ready  to  'cut  loose  from'  a  secure  position  and  assured  income, 
and  who  have  the  gift  of  imparting  their  enthusiasm  to  other  restless  individuals  (technicians, 
salesmen,  laborers)  and  to  still  others,  with  private  capital,  [the  present  writer's  italics]  who 
are  willing  to  take  great  risks  for  the  chance  of  great  gains — it  is  only,  in  a  word,  adventurers 
who  found  new  industries.     The  aviation  industry  has  been  no  exception  •  •  •  " 

3  These  figures  are  as  given  by  the  United  States  Geological  Survey  and  differ  considera- 
bly from  the  figures  of  both  the  census  and  the  National  Electric  Light  Association. 


1919-1929  769 

in  the  cyclical  variations  of  the  period  than  the  immediate  effects  of  the 
investment  in  power  plant,  transmission  lines,  and  distribution,  we  may 
yet  note  that  from  1917  to  1927  balance-sheet  values  of  power  plants 
increased  from  about  3  to  about  9.4  billions1  and  that  more  than  1.5 
billions  of  electrical  stock  and  bonds  was  issued  in  the  yearly  average 
between  1924  and  1930 — the  maximum  of  2,150  millions  occurred  in 
1927 — of  which  perhaps  something  less  than  two-thirds  was  spent  on  new 
construction  and  extension.2  Gross  earnings  of  the  electric  light  and 
power  industry  reached  2.1  billion  dollars  by  1929, 3  when  household 
consumption  was  responsible  for  604  millions,  industrial  and  commercial 
for  about  1.2  billion,  street  lighting  and  traction  for  the  rest. 

Prices,  of  course,  differed  widely,  not  only  locally  but  also  as  between 
customers:  in  1929  the  leather  industry,  for  instance,  paid  $28  for  1,000 
kilowatt-hours  and  the  chemical  industry  5.9,  12.7  dollars  being  the 
average  for  that  year  as  given  by  the  census.  In  the  average,  however, 
they  fell.  The  national  average  price  of  current  used  in  households  is  a 
no  less  doubtful  matter.  The  semiofficial  figures  are  per  kilowatt-hour: 
16.2  cents  around  the  turn  of  the  century,  about  9  cents  for  1912,  roughly 
7.5  at  the  beginning  of  our  period,  during  which  it  slowly  but  steadily 
fell  to  6.3  in  1929,  or  about  3.8  cents  in  terms  of  the  prewar  purchasing 
power  of  the  wage  earners'  dollar.4  This  behavior  of  prices  is  accounted 
for,  on  the  one  hand,  not  only  by  the  actual  or  potential  competition  of 
industrial — as  distinguished  from  "public" — stations,  but  also  by  "com- 
modity competition" — gas,  nonelectrical  motors — and  the  necessity  of 
building  up  new  demand:  the  electrification  of  the  household  and  of  the 
farm  in  particular  was  to  a  large  extent  a  question  of  price.  On  the 
other  hand,  the  growth  of  units  of  control  and  the  establishment  of  local 

1  Census  data;  we  might  also  compare  the  1002  figure  of  investment  in  that  sense — 
about  0.5  billions — to  the  1932  figure — nearly  13  billions,  about  $384  per  kilowatt  installed. 
But  not  much  confidence  can  be  placed  in  any  of  those  data.     Variations  in  the  significance 
of  the  monetary  unit  apart,  the  figures  of  different  census  are  not  strictly  comparable, 
because  they  do  not  include  exactly  the  same  things.     The  above  comparison  can  at  best 
give  an  idea  about  orders  of  magnitude. 

2  About  two  billions  of  the  total  was  raised  between  1922  and  1932  from  customers. 
The — on  the  whole — rising  standing  of  the  industry  drew  new  types  of  investors  toward 
its  mortgage  securities,  as  life  insurance  companies  and  also — compare,  for  example,  the 
New  York  State  Act  of  1927 — savings  banks.     In  its  then  dimensions  this  was  a  postwar 
development.     Including  the  gas  industry,  total  investment  by  all  classes  of  investors  is 
said  to  have  reached  the  figure  of  18  billions  by  1932,  including  what  may  be  termed  th<> 
distress  issues  of  that  year. 

3  The  development  is  interesting:  1902,  85.7  millions;  1907,  175.6;  1912,  302;  1917,  521; 
1922,  1,072;  1929,  2,107.     Cf.  Wall  Street  Journal,  June  28,  1930. 

4  According  to  the  cost-of-living  index  of  the  National  Industrial  Conference  Board. 
The  Department  of  Labor's  prices  of  current  are  somewhat  higher,  but  their  tendency  is 
the  same. 


770  BUSINESS  CYCLES 

and  sectional  monopolies  facilitated  discrimination  and  went  far  toward 
eliminating  price  competition  between  those  units,  while  their  struggles 
were  transferred  to  the  financial  sphere.  That  explains  why  the  weighted 
average  of  prices  did  not,  within  our  period,  fall  correspondingly  to  the 
increase  in  efficiency  of  production,  and  this  again  was  why  most  oper- 
ating companies  were  in  a  position  to  improve  their  financial  status 
considerably1  and  to  weather  the  subsequent  storm  comparatively  well. 
The  competing-down  process  and  its  contribution  to  the  general  picture 
of  the  period,  but  especially  to  the  subsequent  Great  Depression,  took 
under  the  circumstances  a  form  which  was  in  many  respects  peculiar. 
It  asserted  itself  mainly  through  shifts  in  industrial  location — electrical 
development  materially  helping,  for  instance,  in  the  industrialization  of 
the  South — and  much  less  directly  as,  for  instance,  in  the  effect  on  coal. 
But  no  difficulty  arises  in  elaborating  this  aspect. 

Technological  advance  was  much  on  the  same  lines  as  in  Europe. 
Water-power  development  played,  of  course,  a  great  role:  from  1924  to 
1928  it  progressed  at  a  greater  rate  than  the  capacity  of  steam  plants, 
reaching  an  output  of  29  billion  kilowatt -hours  by  1930  though  at  the 
end  of  the  period  steam  began  to  gain  ground  relatively.  The  use  of 
fuel  oil  and  gas  was  an  American  peculiarity.2  Otherwise  we  observe 
the  general  tendencies  toward  larger  capacity  of  stations — the  number  of 
plants  fell  by  one-third  between  1922  and  1929 — and  superpower  zones. 
Since  in  extending  electrical  enterprise  to  foreign  countries  capital  counts 
for  almost  everything,  the  success  of  American  groups,  especially  in 
South  America,  is  easy  to  understand  (American  and  Foreign  Power  Co.). 
About  one  billion  went  to  South  America,  Europe,  Asia,  and  into  what 
presently  turned  out  to  be  so  many  traps. 

Considering  the  technological  nature  of  much  that  was  done,  mergers, 
partly  also  aiming  at  the  control  of  gas  concerns,  were  the  unavoidable 
concomitant  of  this  development.  The  financial  instrument  of  the  hold- 
ing company  lying  ready  at  hand  therefore  experienced  a  new  vogue  of 
unprecedented  dimensions.  Power  finance  definitively  passed  out  of  the 
hands  of  the  manufacturing  industry  and  coordination  resulted  from  a 
struggle  within  the  power-producing  sphere,  in  which  the  groups  men- 
tioned above  emerged  or  conquered.  Since  this  struggle  involved  com- 

1  The  general  question  of  raising  funds  for  consolidation  and  investment — in  part — by 
"taxing"  consumers  tw.  borrowing  is  conveniently  discussed  by  means  of  the  model  of  a 
socialist  state  faced  by  the  analogous  problem.     It  cannot  be  disposed  of  by  an  argument 
on  the  lines  of  Marshall's  concept  of  consumers'  rent  because  this  covers  at  best  the  sta- 
tionary case.     All  that  matters  for  us,  however,  is  the  fact  of  that  price  stability  and  its 
possible  effects  on  the  cyclical  process  under  the  three  headings  of  "rigidity,'*  profits,  and 
financial  status. 

2  The  use  of  natural  gas,  itself  a  major  feature  of  the  period,  increased  in  the  decade 
after  1919  from  21.4  billion  cubic  feet  to  77. 


1919-1929  771 

petitive  bidding  for  strategic  positions,  such  geographical  and  commercial 
rationalization  as  was  achieved  was  accompanied  by  the  growth  of  a 
huge  structure  of  debt — closed  mortgages,  open-end  mortgages,  insured 
debentures — and  share  capital,  which  was  out  of  proportion  with  the 
effects  of  that  rationalization,  and  not  only  provided  food  for  purely 
financial  maneuvers  and  speculative  excesses  of  a  type  suggestive  of  the 
railroad  age,  but  also  jeopardized  the  banking  system,  since  power  securi- 
ties loomed  large  in  its  collateral  and  since  many  leading  banks,  among 
them  the  National  City,  the  Chase  National,  the  Bankers'  Trust,  the 
Guaranty  Trust,  associated  their  fortunes  directly  with  power  enterprise 
and  in  fact  functioned  in  some  cases  as  the  agents  of  ultimate  centraliza- 
tion. Without  going  further  into  this  well-known  matter,  we  will  note 
that  the  great  boom  in  power  finance — and  real  investment — belongs  to 
the  second  half  of  our  decade.  It  was  a  feature  of  the  fourth  Juglar  and 
clearly  basic  to  its  prosperity  phase.  In  fact,  building  construction, 
power  development — together  with  developments  in  other  branches  of 
the  utility  field  which  we  cannot  stay  to  discuss  but  which  also  fit  into 
our  general  idea  of  the  processes  of  a  Kondratieff  downgrade1 — would  in 
themselves  suffice  to  account  for  the  behavior  of  aggregative  time  series 
during  the  period. 

The  major  instances  of  the  propelling  and  dislocating  effects  of  power 
developments  are  obvious,  and  description  of  the  sum  total  of  all  the 
minor  ones  is  impossible  within  this  sketch.  But  it  should  be  emphasized 
in  view  of  popular  dirges  about  lack  of  investment  opportunity  that  the 
work  of  electrification — as  much  of  it  even  as  is  technologically  and 
commercially  possible  at  the  moment  or  in  immediate  prospect — is  not 
nearly  completed.  There  is  enough  investment  opportunity  from  this 
source  alone  for  many  a  cycle  to  come.  Even  industry  is  as  yet  but 
imperfectly  electrified — perhaps  to  something  like  75  per  cent — and  so 
are  households,  while  but  a  beginning  has  been  made  in  the  electrifica- 
tion of  farms  and  of  transportation.  Only  the  telephone  and  electric 
lighting  can  reasonably  be  said  to  have  exhausted,  ex  visu  of  present 
technology,  the  bulk  of  their  possibilities,2  although  the  automatic  tele- 

1  Utility  developments  form  part  of  the  picture  which  we  expect  a  Kondratieff  down- 
grade to  reveal,  because  they  are  to  a  large  extent  a  function  of  real  income  and  its  rate 
of  change.     Accordingly,  we  find  expansion — induced  expansion — in  the  utility  field  in 
the  two  last  decades  of  the  first  Kondratieff  as  well  as  in  the  downgrade  of  the  second 
(eighties  and  early  nineties).     We  find  the  same  phenomenon  in  the  present  instance. 

2  A  criterion  of  whether  or  not  an  industry  is  past  its  first  spurt  is  in  some  cases  afforded 
by  the  presence  or  absence  of  setbacks  in  its  expansion.     As  far  as  the  writer  is  able  to 
make  out,  the  number  of  telephones  installed  increased  without  any  break,  depression  or 
no  depression,  from  1876,  when  there  was  no  commercial  installation,  to  1980  inclusive, 
after  which  year  there  was  a  fall.     Recovery  set  in  in  1934,  but  the  precrisis  figure  was  not 
nearly  attained  in  1935.     However,  within  our  period  the  number  of  telephones  installed 


772  BUSINESS  CYCLES 

phone — installation  was  zero  in  1892  and  only  1.7  per  cent  of  the  total  of 
telephones  installed  in  1919,  but  nearly  32  per  cent  in  1930 — which  must 
be  listed  among  the  innovations  of  the  period  under  discussion,  affords 
a  good  illustration,  if  one  be  needed,  for  the  fact  that  even  perfect  satura- 
tion of  existing  demand  need  not  call  a  halt  of  "progress." 

Production  of  electrical  equipment  had,  ever  since  1915,  increased  at 
a  greater  rate  than  production  of  power  and  continued  to  do  so  until 
1929.  Its  value  was  about  1  billion  dollars  in  1919  and  nearly  2.5  billions 
in  1929. 1  Examples  of  new  industries — and  the  "diversifying"  effect  of 
power  production — abound.  We  will  merely  note  that  the  spectacular 
expansion  of  the  radio  and  the  refrigerator  industries  dates  from  1926. 
The  quarter  of  a  million  socket  radios  then  in  use  increased  to  over 
7  millions,  the  315,000  refrigerators  to  1,680,000  in  1929.2  Though 
typical  instances  of  downgrade  developments,  these  were  practically 
new  industries  with  histories  of  their  own.  But  they  were  not  so  inde- 
pendent of  the  older  concerns  in  the  industry  of  electric  manufacture  as, 
say,  aviation  is  of  the  automobile  industry.  Generally  speaking,  these 
older  concerns  maintained  their  position  well,  and  proved  in  this  as  in 
other  countries  successful  shells  of  incessant  innovation,  especially  in 
the  heavy-current  field  (General  Electric,  Westinghouse) .  Dollar  volume 
of  output  in  electric  manufacturing  increased  about  sevenfold  between 
1914  and  1929,  arid  about  26  times  from  1899,  the  census  year  nearest 
to  the  beginning  of  the  Kondratieff,  to  1929.3 

b.  The  automobile  industry  led  in  every  upswing  and  out  of  every 
downswing  throughout  the  period,  in  fact  beyond  it,  and  continued  in 
the  Kondratieff  recession  to  qualify  as  well  for  the  role  of  standard 
example  for  the  processes  embodied  in  our  model  as  it  had  done  in  the 
upswing.  Employment  in  motor-vehicle  factories,  not  including  pro- 
duction of  parts,  tires,  and  bodies,  increased  from  about  253,000  in 
1922  to  427,500  in  1929,  the  corresponding  wage  bill  from  about  396  to 
about  775.5  million  dollars.  Passenger-car  registration  as  of  Dec.  31 
increased  without  any  break  from  the  beginning  of  the  series  (1895:4) 

rose  from  about  12.7  millions  in  1919  to  over  20  millions  in  1929.  This  increase  was 
sufficient  to  raise  the  number  of  telephone  operators  by  about  80  per  cent,  in  spite  of  the 
labor-saving  effects  of  the  automatic  telephone  and  some  "  taylorizing." 

1  Census  figures. 

2  Figures  of  the  Edison  Institute.     The  number  of  socket  radios  continued  to  increase 
throughout  the  depression  to  nearly  20  millions  in  1935,  an  example  for  those  initial  spurts 
which  are  impervious  to  depressions.     As  the  reader  will  remember,  such  behavior  is,  if 
anything,  normal  from  the  standpoint  of  our  analysis,  though  in  practice  it  is  not  the 
general  rule.     Similarly,  the  number  of  refrigerators  in  occupied  homes  kept  on  increasing 
without  a  break  and  reached  the  figure  of  7.25  millions  in  1935. 

8  The  number  of  workmen  employed  rose  from  42,000  to  329,000  between  the  census  of 
1899  and  1929. 


1919-1929  773 

to  1929  (23,121,589),  though  of  course  at  a  decreasing  percentage  rate, 
depressions  affecting  the  latter  only.1  Even  in  the  world  crisis  and  in 
the  year  of  minimum  registration  (1933)  the  total  automobile  retail  and 
service  business,  including  accessories,  filling  stations,  garages,  and  also 
retail  sales  by  wholesalers,  figures  out  at  $4,831,800,000.2  Over  1.1  mil- 
lion persons  were  engaged  in  distribution  and  servicing,  among  them 
756,000  employees  (part-time  included),  receiving  wages  and  salaries 
amounting  to  801  millions.  Quantitative  expansion  and  qualitative 
improvement,  falling  costs,  prices,  and  rates  of  profit  are  obviously  the 
expected  as  well  as  the  actual  characteristics  of  this  industry's  history 
during  our  decade.  However,  since  there  is  no  satisfactory  way  of 
measuring  qualitative  improvement,  and  since  there  was  an  almost 
uninterrupted  shift  from  larger,  heavier,  and  dearer  to  smaller,  lighter, 
and  cheaper  cars — in  1903,  for  instance,  4.2  per  cent  of  automobiles 
produced  cost  $675  and  less,  in  1924  nearly  60  per  cent3 — even  quantita- 
tive expansion  becomes  elusive,  while  indices  of  quoted  prices,  which 
should  moreover  be  corrected  for  variations  in  the  allowances  made  for 
old  cars  "traded  in"  and  for  other  forms  of  rebates,  cannot  indicate 
more  than  a  tendency  which,  of  course,  they  understate.4  From  1916  on, 
profits  of  individual  firms  not  only  fell  but  also  became  more  nearly  equal. 

1  Tax-exempt  official  cars  are  excluded.     Registration  of  trucks  increased  through 
1980.     Possibly  a  "break  in  trend"  in  registration  of  passenger  cars  may  be  said  to  have 
occurred  about  1929,  after  which  it  may  be  expected  to  move  over  time  roughly  as  popula- 
tion in  the  age  groups  between  25  and  60. 

2  United  States  Census  Bureau,  "Census  of  American  Business  for  1933."     It  is  of 
some  interest  to  note  that  in  the  same  year  state  highway  expenditure  was  about  666  million 
dollars  (United  States  Bureau  of  Public  Roads),  total  rural  highway  expenditure  about 
1.5  billions. 

3  See  Epstein,  op.  cit.t  p.  336.     Needless  to  say  that  does  not  measure  that  shift  with 
any  exactness. 

4  On  this  point  see  Professor  Epstein,  op.  cit.,  p.  47.     He  tries  to  convey  an  idea  by 
giving  instances  of  the  selling  prices  of  "fairly  comparable  models,"  for  instance,  of  a 
Packard  which  sold  at  $7,000  (without  equipment)  in  1904  and  at  $2,585  (equipped)  in 
1924.     The  Bureau  of  Labor  Statistics  wholesale  index  of  automobile  prices  falls  sharply  to 
1916,  then  rises  to  a  peak  in  1920,  from  which  it  again  sharply  falls  to  1926 — eight  points 
below  the  minimum  of  1916.     Then  it  rises  to  1929,  partly  at  least  owing  to  the  reconstruc- 
tion of  the  Ford  works.     Another  index  has  been  constructed  by  Mr.  J.  W.  Scoville, 
Behavior  of  the   Automobile   Industry   in   Depression,    (Econometric   Society   address, 
December  1935;  published  separately),  which  also  displays  a  trough  in  1916.     The  subse- 
quent peak,  however,  comes  in  1918.     From  it  this  index  falls  precipitously  to  1923,  to 
rise  to  and  reach  a  peak  in  1927,  after  which  there  is  decline,  much  stronger  than  in  the 
case  of  the  Bureau  of  Labor  Statistics  index,  to  1933.     Neither  index,  of  course,  overcomes 
the  fundamental  difficulties,  which  also  vitiate  cost  figures.     Attempts  have  been  made  to 
arrive  at  a  more  telling  picture  by  computing  prices  of  cars,  or  of  automotive  products  in 
general,  per  pound.     The  Automobile  Manufacturers'  Association  constructed  an  index 
based  on  the  average  list  prices  of  the  lowest  priced  five-passenger  closed  model  of  each 
make,  weighted  by  its  relative  share  in  new-car  registration.     This  index  overstates  the 


774  BUSINESS  CYCLES 

The  industry  did  not  simply  expand  in  function  of  the  increase  in  real 
income  but  helped  to  bring  it  about.  The  former  nexus,  however, 
steadily  gained  in  importance  at  the  expense  of  the  latter,  as  had  been 
the  case  with  cotton  after  the  Napoleonic  wars  and  with  railroads  from 
the  eighties  on.  Innovations,  increasing  in  number  while  individually 
decreasing  in  importance,  are  typically  of  the  downgrade  type.  From 
1912  on,  designs  became  more  stable.  Considerable  progress  in  the 
standardization  of  parts  and  in  the  rationalization  of  assembling  reduced 
costs  as  did  progress  in  subsidiary  industries — tires,  nitrocellulose  lac- 
quers and  fast-drying  solvents,  and  so  on.  Equally  important  or  more 
so  were  the  changes  in  organization  and  financing  that  were  in  part 
induced  by  the  struggle  for  survival  within  the  industry,  in  which  inces- 
sant innovating  and  expanding  into  the  low-price  market  was  a  matter 
of  life  and  death.  Compet ing-down  went  on  at  a  rapid  rate.  The  rise 
in  price  level  after  1916  helped  to  keep  failures  and  exits  at  a  low  and 
decreasing  figure,  and  even  the  setback  of  1918,  when  both  production 
and  wholesale  value  fell  absolutely  for  the  first  time,  cost  few  lives. 
But  after  1921,  when  production  and  wholesale  value  again  fell  abso- 
lutely, exits — not  necessarily  failures — increased  sharply  in  the  midst  of 
spectacular  expansion  of  the  industry  as  a  whole,  reaching  21  per  cent 
in  1924.  In  1923  and  1924  no  less  than  29  firms  went  out  of  business, 
17  of  them  war  and  postwar  foundations.  Of  the  101  plants — makers, 
not  concerns — whose  annual  production  of  passenger  cars  was  5,000  or 
less  in  1920  only  11  survived  in  1930;  of  the  23  whose  annual  production 
was  from  5,000  to  25,000,  also  11;  while  we  still  find  all  of  the  10  which 
produced  over  25,000  in  1920.1  By  1918,  70  per  cent  of  all  automobiles 
produced  in  this  country  and  Canada  came  from  the  three  largest  pro- 
ducers, by  1921  80  per  cent,  and  by  1935  nearly  90  per  cent. 

Considering  that  the  car  of  the  masses2  became  a  reality,  while  the 
industry,  which  had  always  been  monopolistically  competitive,  developed 

case — it  falls  by  40  per  cent  from  1925  to  1932,  see  chart  on  p.  6  of  Mr.  Sloane's  message 
to  the  General  Motors  shareholders  of  Dec.  31,  1937 — but  is,  in  the  particular  circum- 
stances, perhaps  somewhat  less  objectionable  than  it  would  be  in  general.  For  the  behavior 
or  profits,  see  Epstein,  op.  cit.,  pp.  243,  256,  264.  There  it  will  be  seen  that  the  "bonanza 
period"  lasted  well  into  the  Kondratieff  recession  (1916,  according  to  Professor  Epstein), 
which  conforms  to  expectation.  It  is,  however,  interesting  to  observe  how  the  leading 
firm  fared.  The  ratio  of  profits  to  net  worth  in  the  Ford  concern  fell  from  its  1907  peak 
with  well-marked  cyclical  fluctuations  heading  toward  zero  until  the  plant  underwent  its 
complete  reconstruction — which  amounted  to  a  second,  though  induced,  innovation — in 
1927.  Number  of  new  companies  founded  followed  those  fluctuations  with  an  average 
lag  of  between  one  and  two  years. 

1  J.  W.  Scoville,  op.  cit.,  p.  24. 

2  Tt  should  be  mentioned  that  the  "automobilization"  of  the  farm  went  further  than  did 
its  electrification.     At  the  end  of  our  period  about  60  per  cent  of  all  farms  were  equipped 
with  motor  vehicles,  and  about  one-quarter  of  all  trucks  were  in  farm  use. 


1919-1929  775 

a  typically  oligopolistic  situation,  we  cannot  help  being  painfully  aware 
once  more  of  the  somewhat  less  than  realistic  character  of  the  general 
conclusions  arrived  at  by  the  leading  theorists  of  monopolistic  competi- 
tion.1 In  fact,  it  should  be  obvious  that  the  behavior  of  the  motorcar 
industry  during  our  decade  could  be  described  much  more  convincingly 
in  terms  of  perfect  competition  working  under  the  conditions  of  a  new 
industry  in  the  course  of  being  absorbed  by  or  inserted  into  the  economic 
system.  In  the  course  of  this  development,  ever  since  about  1916, 
methods  of  financing  changed  significantly.  "Outside  capital"  began 
to  play  a  greater  role.  We  need,  however,  only  mention  the  direct  con- 
tact established  by  General  Motors  with  the  open  market  and  its  policy 
— followed  by  the  other  concerns — of  financing  the  consumer.  Never- 
theless, owned  capital  accumulated  from  profits  and  retailers'  and  fur- 
nishers' credit  remained  the  industry's  most  important  sources  of  means, 
and  this  accounts  for  much  which  strikes  the  observer  as  particularly 
"sound"  about  it.  Net  tangible  assets  of  motor-vehicle  manufacturing 
plants  reached  their  maximum  of  about  2.1  billion  dollars  in  1926  and 
then  steadily  fell,  though  up  to  the  crisis  but  slowly.  However  unreliable 
any  inference  from  this  may  be,  it  seems  clear  that,  barring  the  Ford 
plant,  the  great  wave  of  investment  belongs  to  the  third  and  not  to  the 
fourth  Juglar. 

In  order  to  prove  with  quantitative  precision  how  much  of  the  proc- 
esses of  the  period  and  of  the  behavior  of  aggregates  can  be  explained  by 
the  motorcar  developments  alone,  it  would  be  necessary  to  go  fully  into 
what  they  meant  for  the  steel,  copper,  and  equipment  industries  and 
so  on.2  We  will,  however,  confine  ourselves  to  one  remark  on  the 
petroleum  and  another  on  the  rubber  industry.  Innovations  that  have 
already  been  mentioned  (Chap.  VII:  flooding,  cracking,  hydrogenation, 
extension  of  new  uses  such  as  fueling  of  locomotives  and  ships,  by-prod- 
ucts) and  the  discovery  and  development  of  new  oil  fields  account  for 
the  fall  in  gasoline  prices  (excluding  tax)  from  $0.2411  per  gallon  in 
1919  to  $0.1557  in  1929  and — gasoline  consumption  did  not  fall  until 

1  Some  exponents  of  the  practical  implications  of  those  conclusions — restriction  of 
output  at  rising  prices  and  falling  profits,  uneconomically  undersized  plants,  and  so  on — • 
are,  realizing  that  there  is  danger  that  those  conclusions  might  look  to  us  like  caricatures, 
in  the  habit  of  listing  the  motorcar  industry  as  an  exception.     But  it  is  only  the  outstanding 
example  of  a  very  large  class.     The  tire  industry,  to  be  mentioned  presently,  is  another. 
Compare  what  has  been  said  on  these  points  in  Chaps.  II  and  X. 

2  Even  the  railroads,  which  on  balance,  of  course,  suffered,  did  not  go  without  a  share. 
Transport  of  vehicles  themselves,  of  steel,  gasoline,  and  lubricating  oil  contributed  a  non- 
negligible  amount  to  their  revenue.     This  amount  becomes  even  important  if  we  include 
road-building  material,  although  the  Automobile  Manufacturers'  Association's  estimate 
of  over  265  millions  for  1934  seems  somewhat  optimistic.     Net  premia  of  all  types  of 
automobile  insurance  amounted  to  nearly  411  millions  in  the  same  year,  according  to  the 
Insurance  Year  Book  Service. 


776  BUSINESS  CYCLES 

1932 — $0.1178  in  1931, l  which  shows  that  the  petroleum  industry  was 
not  passively  drawn  along  by  the  growth  of  demand.  Yet  it  comes 
sufficiently  near  to  this  pattern  to  qualify  as  an  instance.  This  is  par- 
ticularly evident  at  the  beginning  of  the  period.  In  1920  prices  of  oil 
and  gasoline  rose  considerably  (peak  of  the  period),  so  much  so  as  to 
throw  them  out  of  line  with  those  of  competing  fuels  and  as  to  restrict 
the  use  of  fuel  oil  by  railroads — the  Great  Northern,  for  instance,  con- 
verted 70  locomotives  into  coal  burners.  This  followed  upon  the  doubling 
of  automobile  production  in  1920  as  compared  with  1918,  with  which  the 
gasoline  production  was  then  unable  to  keep  pace.  An  oil  boom  started 
accordingly,  which  almost  coincided  with  deep  depression  in  other  lines. 
Issues  of  oil  securities  were  at  a  peak  early  in  1920  and  again  toward  the 
end  of  the  year  and  at  the  beginning  of  1921.  It  is  worth  while  to  men- 
tion that  the  only  cities  in  the  country  which  experienced  greater  building 
activity  in  November  1920  than  in  November  1919  were  Los  Angeles, 
Baltimore,  and  New  Orleans,  and  that  the  Californian  cities  all  showed 
large  gains  in  their  clearing  figures  while  these  declined  in  the  rest  of 
the  country.  At  the  beginning  of  1921  there  was  a  large  oil  merger 
(Barnsdall  Corporation).  Further  developments  followed  and  crude 
prices  reacted  promptly,  Midwestern  prices,  for  instance,  falling  to  $1 
a  barrel  in  the  summer  of  1921,  as  compared  with  $3  in  January.  We 
need  not,  however,  follow  the  history  of  this,  in  many  respects,  peculiar 
case.2 

The  rubber  industry  was,  of  course,  also  "drawn  along."  But  its 
own  innovations  were  much  more  in  evidence.  As  we  have  seen  else- 
where, beginnings  date  far  back  (Goodyear  vulcanization  to  1839,  for 
instance)  or  at  any  rate  to  the  Kondratieff  prosperity  (reclaiming,  e.g., 
1899,  acceleration  of  the  vulcanizing  process  1906;  but  commercial  suc- 
cess of  synthetic  rubber  came  after  our  period),  the  use  of  various  pig- 
ments in  order  to  increase  the  durability  of  rubber  compounds  (1916) 
being  the  only  "inventive"  innovation  of  the  twenties.  It  was  again  the 
"spreading"  by  means  of  discovering  new  and  developing  old  industrial 
uses  for  rubber  (flooring,  rubber  cushions,  rubber  linings,  mountings, 
bumpers,  and  so  on)  which  was  a  feature  of  the  period  under  discussion. 
In  the  field  of  the  most  important  article  the  great  new  thing — though 
also  invented  long  ago  (R.  W.  Thompson,  patented  1845) — was,  of 
course,  the  pneumatic  tire  (1916),  which  followed  upon  the  success  of 

1  Figures  of  the  American  Petroleum  Industries  Committee;  tank  wagon;  50  selected 
cities. 

2  We  may  note  that  in  1927  the  industry  employed  1.25  million  people  and  that  equip- 
ment facilities,  including  pipe  lines,  were  valued  at  about  11  billions.     There  were  then 
830  refineries  with  a  capacity  of  8  million  barrels  crude  a  day,  while  the  818,600  wells 
produced  2,4  millions. 


1919-1929  777 

the  cord  and  may  be  said  to  have  imparted  immediately  a  significant 
impulse  to  long-distance  trucking,1  although  at  as  late  a  date  as  July  1, 
1920,  the  India  Rubber  World  (p.  633)  professed  itself  unable  to  believe 
that  the  solid  tire,  which  had  greatly  improved  in  reaction  to  the  intru- 
sion of  the  new  competitor,  would  be  crowded  out.  At  the  same  time  the 
commercial  opportunity  for  low-pressure  tires  for  passenger  cars  mani- 
fested itself  in  the  habit  of  many  motorists  to  underinflate  their  tires 
for  the  sake  of  comfort.2  By  1923,  21  companies,  among  them  prac- 
tically all  the  leaders  of  the  trade,  were  making  such  tires,  experimentally 
or  commercially,  and  several  automobile  manufacturers  had  adopted 
them  as  part  of  the  regular  equipment  of  their  cars,  while  others  listed 
them  as  optional.  A  "revolution"  in  tire  making,  the  more  important 
because  it  involved  considerable  new  investment,  announced  itself. 
There  was  still  resistance  to  overcome.  But  improvement  and  stand- 
ardization— as  to  rim  requirements — carried  the  innovation  suddenly  to 
definitive  success  about  1925,  after  one  of  the  tire  companies  had  taken 
the  bold  step — in  the  midst  of  doubts  about  practicability  and  the  prob- 
able reaction  of  the  public — to  bring  out  balloon  tires  for  all  standard 
rims  and  thus  to  make  a  bid  for  immediate  replacement  of  practically 
all  tires  in  use.  The  aspect  of  the  market  changed  within  a  few  months, 
and  the  "host"  followed  the  innovator  promptly.  There  is  no  need  of 
going  into  the  illustrative  virtues  of  the  case  or  the  quantitative  impor- 
tance of  it  for  the  fourth  Juglar.3  With  quick  changes  in  production 
functions,  the  competing-down  process  asserted  itself  strongly.  We 
shall  interpret  in  this  sense  the  symptoms  of  overinvestment  and  over- 
production,4 observable  already  in  1923  and  again  after  1926,  and  expect 
a  contribution  to  the  picture  of  the  subsequent  crisis  from  this  industry. 

1  Registration  of  trucks  increased  from  136,000  in  1915  to  over  2  millions  in  1924,  much 
more  than  passenger-car  registration.     This  would  hardly  have  been  possible  without 
the  success  of  the  pneumatic  tire,  first  displayed  by  the  Goodyear  Wingfoot  Express. 

2  This  instance  shows  the  relation  between  "wants"  and  enterprise  (in  our  sense)  very 
nicely.     Consumers  by  thus  clearly  indicating  their  wishes  displayed  an  altogether  unusual 
amount  of  initiative,  which  greatly  facilitated  the  entrepreneurial  act.     Yet  they  did  not, 
off  their  own  bat,  "demand"  balloons  and  that  practice  of  underinflating  might  have  gone 
on  indefinitely  without  giving  rise  to  what  almost  amounted  to  a  new  industry.     This 
development  was,  no  doubt,  conditioned,  but  it  was  as  little  brought  about  by  consumers' 
behavior  as  was  the  development  from  fabrics  to  cords,  supertwists,  and  rayon  cords. 

3  Employment  in  the  industry  expanded  considerably  at  wage  rates  which  head  the 
list  of  basic  industries  and  were  above  the  average  of  all  manufacturing  industries.     It 
has  been  stated  that  in  1908  a  workman  in  tire  manufacture  received  on  the  average  40 
cents  per  hour  and  could  have  bought  for  $35  a  tire  which  was  good  for  about  2,000  miles, 
while  in  1936  he  received  88  cents  and  was  able  to  buy  a  tire  which  cost  $8  and  gave  20,000 
miles  of  service,  so  that  an  hour's  labor  gave  him  about  95  times  as  much  tire  service  in 
1936  as  in  1908. 

4  E.  G.  Nourse  and  Associates,  America's  Capacity  to  Produce,  1934,  pp.  236  and  237, 
estimate  average  utilization  of  capacity  for  1925-1929  at  85.3.     This  is  high  under  the 


778  BUSINESS  CYCLES 

c.  The  heavy  chemical  industry  had,  as  we  have  seen,  developed  well 
before  the  war,  but  enterprise  in  the  organic  branch  was  entirely  condi- 
tioned by  the  seizure  of  German  patents  and  later  on  by  protection.1 
Prices  of  chemicals,  which  according  to  the  B.  L.  S.  index  (1926  =  100) 
were  at  89.4  in  1913  and  which  had,  owing  to  the  practical  cessation  of 
German  imports,  soared  to  197  by  1916,  testified  to  the  vigor  of  entre- 
preneurial response  to  those  new  conditions  by  falling  to  97.2  in  1922.2 
Both  the  coal-tar  group  in  all  its  stages,  particularly  in  the  production 
of  dyes,  and  the  aliphatic  group  scored  a  series  of  successes  that  extended 
over  the  whole  of  our  period  and  throughout  the  subsequent  depression 
and  amounted  to  the  creation  of  new  industries.  Investment,  employ- 
ment, wage  bill,  profits,  and  dollar  volume  of  sales — about  2J4  billions 
towards  the  end  of  our  decade — increased  to  a  peak  in  1929  for  the  chemi- 
cal industry  as  a  whole.  Sales  in  the  non-coal-tar  group  continued  to 
increase  without  break  afterward.  Medicinals,  solvents,  perfumes, 
antifreezes,  carbon  tetrachloride,  acetic  anhydride,  camphor,  resins, 
nitrates  (synthetic  iodine  and  synthetic  rubber  came  early  in  the  thirties) 
may  serve  as  examples.  Analysis  of  the  individual  cases  would  show 
little  more  than  so  many  instances  of  the  way  in  which  innovation  works. 

Three  points  only  call  for  additional  comment.  First,  for  the  same 
reasons  as  elsewhere  concentration  of  control  and  research  and  coordi- 
nation of  specialized  large-scale  plants  were  in  evidence  in  this  industry. 
The  Dupont  concern,  like  the  J.  G.  Farben,  expanded  far  beyond  the 
chemical  field.3  The  other  giant,  the  Allied  Chemical  and  Dye  Corpo- 
ration, was  the  result  of  a  merger  in  1920  of  five  big  concerns — three  of 
which  were  in  the  *' heavy"  field — which  to  a  large  extent  complemented 
each  other.  Second,  new  branches  of  industry  emerged  around  what 

circumstances  and  reflects  the  increase  in  production  from  about  2  million  tires  in  1918  to 
nearly  8  million  in  1929.  But  it  is  perfectly  compatible  with  considerable  underutilization 
and  overproduction  (the  latter  in  the  sense  that  "old"  firms  often  failed  to  cover  their 
costs)  in  spots. 

1  Developments  illustrate,  however,  the  rationale  of  the  distinction  between  condition- 
ing innovation  and  innovation  itself.     Whoever  feels  inclined  to  doubt  it  is  invited  to  try 
to  deduce  these  developments  uniquely  and  ceteris  paribus  from  those  two  events. 

2  Then  they  rose  to  hover  around  100  (there  was  an  upward  movement  in  1925)  through 
1928,  when  they  embarked  in  consequence  of  quantitative  expansion  at  falling  costs  on  another 
downward  course.     It  is  important  to  note,  first,  that  this  again  affords  an  illustration  of 
popular  statements  about  rigidity  and,  second,  how  much  more  obvious  is  the  influence  of 
individual  prices,  falling  under  the  impulse  of  innovation,  on  the  price  level  than  any 
influence  of  autonomous  monetary  factors  acting  through  the  latter  on  individual  prices. 
There  should  be  no  uncertainty  about  the  diagnosis  and  the  significance  of  such  facts 
that  say  indigo  fell  from  $2  a  pound  in  1916  to  14  cents  in  1982,  or  barbital  and  veronal  to 
about  one-seventh  and  aspirin  to  about  one-eighth  of  its  prewar  price. 

3  The  interest  taken  in  General  Motors  is  one  example  of  this  method  of  acquiring  and 
buttressing  markets. 


1919-1929  779 

may  be  called  the  production  of  chemical  fundamentals.  A  host  of  small 
and  medium -sized  firms  took  up  the  production  of  a  truly  unsurveyable 
variety  of  drugs,  cosmetic  articles,  and  so  on.  The  results,  as  distin- 
guished from  the  formal  properties,  of  monopolistically  competitive  situ- 
ations are  much  more  in  evidence  in  this  group  and  within  its  army  of 
retailers  and  advertisers  than  among  the  few  big  producers  of  the  basic 
stuffs.  For  us  it  is  important  to  note  the  quantitative  importance  of  this 
trade  and  to  account  for  its  spectacular  expansion:  the  unrivaled  oppor- 
tunity which  it  exploited  was  one  of  the  consequences  of  the  increase  in 
the  real  income  of  the  masses  which  left  even  the  lowest  income  groups 
with  a  surplus  that  was  not  a  priori  allocated  to  specific  purposes  but 
ready  to  go  wherever  advertisements  beckoned.  The  phenomenon  thus 
fits  well  into  our  ideas  about  downgrade  developments.  Third,  the 
chemical  industry  displays  the  (secondary)  competing-down  process 
within  the  innovating  line  much  less  than,  say,  the  automobile  industry; 
but  it  displays  the  (primary)  competition,  i.e.,  the  competition  with  other 
commodities  or  older  methods  of  producing  the  same  commodities  much 
more  than  almost  any  branch  of  economic  activity.  In  some  cases  its 
innovations  act  through  other  spheres  of  production,  agriculture  for 
instance.  In  others  they  act  directly  and  then  with  a  promptness  to 
the  consequences  of  which  the  social  fabric  of  capitalism  may  well  prove 
unequal  some  day.  Chemistry  provides  not  only  acceptable  and  cheap 
substitutes  for  things  that  are  the  basis  of  much  employment  and  invest- 
ment, but  quite  often  exactly  the  very  same  things — frequently  in  a  better, 
especially  more  uniform  and  more  reliable  quality,  as  for  instance  in  the 
case  of  varnishes  and  dyes — which  had  been  produced  by  nonchemical 
methods  before.  It  does  so  almost  always  at  a  cost  which  eventually, 
though  not  as  a  rule  immediately,  falls  far  below  the  level  attainable  by 
the  latter.  In  such  cases  large  sectors  of  the  economic  organism  may 
have  to  go  out  of  operation  at  very  short  notice.  If  the  consequences 
have  so  far  not  made  themselves  more  strongly  felt  in  our  three  countries, 
this  is  because  they  mostly  impinged  on  others,  on  Chile  for  instance  or 
India — or  in  the  case  of  madder,  on  the  countries  from  Southern  France 
to  Asia  Minor — or  Sicily  (citric  acid,  1927)  or  pro  futuro  on  the  rubber- 
producing  countries.  The  United  States,  England,  and  Germany  were, 
during  the  period  under  survey,  not  much  affected  in  this  respect,  and 
whatever  effect  there  was  was  rather  favorable.  But  more  serious  dis- 
locations may  arise  from  such  developments  some  of  which  are  obviously 
imminent.  The  term  revolution  acquires  in  this  connection  a  particu- 
larly ominous  meaning.  Depressive  influences  may  emanate  from  this 
line  of  advance  by  comparison  with  which  anything  that  can  be  effected 
by  action  on  monetary  aggregates,  central  bank  action  included,  is  of 
negligible  importance. 


780  BUSINESS  CYCLES 

The  rayon  industry,  of  course,  owed  much  to  the  tariff,  and  its  great 
concerns  owed  much  to  their  control  of  patents.  But  in  all  other  respects 
the  case  is  strikingly  analogous  to  that  of  the  automobile  industry.  We 
have  a  sharp  competitive  struggle  at  the  beginning  of  the  period,  partly 
due  to  the  numerous  short-lived  foundations  after  the  war,  from  which, 
as  has  been  stated  elsewhere,  emerged  three  concerns  which  accounted 
for  about  90  per  cent  of  the  production  of  the  country.  In  this  oligo- 
polistic setup  the  great  expansion  of  consumption  took  place1  which  was 
but  little  affected  by  the  world  crisis.  Wholesale  prices  (150%denier,  A 
grade,  New  York)  fell  from  the  1918  peak  to  about  the  prewar  level 
(1914,  $1.96  per  pound)  in  1925,  and  were  at  $1.25  by  1929.2  Profits  per 
pound  of  product  steadily  declined,  although  in  the  case  of  the  American 
Viscose,  which  remained  the  leading  producer  throughout,  they  were 
still  58  cents  in  1928.3  Other  textiles,  though  not  without  some  propelling 
influences — higher  cotton  consumption  per  spindle,  production  of  cord, 
artificial  leather,  broadcloth  shirting,  fancy  woolens  and  so  on — behaved 
like  the  old  industries  they  were.  Quantitative  expansion  and  qualita- 
tive improvement  were  considerable,  and  there  was  much  rationalization 
in  details.  This  does  not  alter  the  fundamental  traits  of  the  picture, 
which  are  reflected  in  the  behavior  of  prices.4  Continuing  locational 
shifts  caused  as  much  sectional  depression  as  sectional  prosperity.  The 
Department  of  Labor's  combined  index  of  employment  kept  steady 
throughout  the  decade,  but  nevertheless  marks  strongly  the  upswing 
that  set  in  during  the  second  half  of  1925. 

d.  As  far  as  changes  in  production  functions  go,  the  iron  and  steel 
industry  should  really  be  dealt  with — as  should  metallurgy  in  general — 
in  connection  with  chemistry  and  electricity.  It  suffices  to  mention 
the  career  of  light  alloys,  the  first  stage  of  which  was  run  during  our 
decade,  especially  from  1925.  There  were  also  technological  and  organi- 
zational changes  of  other  types,  such  as  continuous  rolling  or  the  crowding 
out  of  the  merchant  furnace,  and  of  course  many  improvements  of  the 

1  Deliveries — shipments  of  American  firms  plus  imports  minus  exports — as  computed 
by  the  Textile  Economics  Bureau  and  published  in  the  Rayon  Organon  and  elsewhere, 
increased  nearly  threefold  from  1923  to  1927  and  again  by  about  50  per  cent  to  1931.     They 
do  not  include  acetate  rayon,  which  gained  ground  steadily  and  was  in  1925  little  more 
than  3  per  cent  of  the  total  production,  but  7  per  cent  by  1929  (nearly  22  per  cent  in  1935). 

2  What  has  so  far  been  the  minimum,  50  cents,  was  reached  in  April  1933.     Here  again 
the  contribution  to  the  price  level  of  the  crisis  period  is  more  significant  than  the  contribu- 
tion of  that  level. 

8  The  oligopolistic  pattern  developed  into  a  monopoloid  one  through  the  merger  in 
1925  (American  Rayon  Production  Corporation). 

4  The  B.  L.  S.  group  index  for  cotton  goods  (1926  =  100)  was  56  in  1914,  147.5  in  1919, 
and  190.7  in  1920.  Then  it  fell  abruptly,  but  was  still  98.8  in  1929.  Only  in  1932  it  fell 
below  the  1914  figure.  The  group  index  for  woolens  and  worsteds  behaved  similarly. 


1919-1929  781 

rationalization  kind  in  individual  lines  or  concerns.  Increasing  use  of 
scrap  in  the  steelmaking  as  well  as  in  the  copper,  aluminum,  and  other 
industries  deserves  particular  emphasis.1  Speaking  broadly,  however, 
the  steel  industry  suffered  in  depression — especially  in  1920  and  1921 — 
and  prospered  in  booms — the  peaks  in  pig-iron  output  occur  in  1923, 
1925,  and  1929 — in  consequence  of  the  general  business  situations,  rather 
than  in  consequence  of  its  own  enterprise.  The  behavior  of  its  prices2 
accords  with  this  impression.  Steel  consumption  increased  strongly, 
however,  in  spite  of  all  the  steel-saving  rationalizations,  which  were 
more  than  offset  by  the  conquest  of  new  uses — steel  increasingly  became 
a  consumers'  good — and  general  expansion.  Per  head  of  population  it 
was,  at  the  end  of  our  period,  seven  times  as  great  as  it  had  been  in  1900. 
Nothing  fundamentally  new  happened  in  the  aluminum  and  copper 
industries.  We  have,  however,  already  observed  in  an  earlier  place  that 
the  former  displayed  great  initiative  in  discovering  and  conquering  new 
ground.  Its  quantitative  expansion — domestic  primary  production 
more  than  doubled  between  1915  and  1929,  while  domestic  consumption 
increased  more  than  threefold,  secondary  production  accounting  for  the 
greater  part  of  the  difference — was  one  of  the  major  industrial  features  of 
the  period,  another  good  instance  of  a  downgrade  development.  Price 
behavior  was  in  accordance  with  this  and  characteristically  different 
from  that  of  finished  steel.  The  absence,  comparatively  speaking,  of 
fluctuations  around  the  fundamental  contour  affords  an  interesting 
example  of  what  "  control "  by  one  firm  really  means  under  conditions 
of  rapid  growth  and  of  commodity  competition.  The  domestic  price 
of  new  aluminum  ingot  99  per  cent  pure  reached  its  war  maximum  by 
1916  and  its  prewar  level  or,  corrected  by  the  B.L.S.  index  of  wholesale 
prices,  a  figure  nearly  30  per  cent  below  it,  by  1922.  It  then  slowly  rose 
to  1925,  afterward  fell  somewhat,  and  was  maintained  at  23.3  cents 

1  This  material-saving  practice,  a  typical  downgrade  development  and   responsible 
for  the  increasing  divergence  between  pig-iron  and  steel  outputs,  of  course  exerts  a  depress- 
ing influence  on  the  production  of  a  number  of  important  raw  materials  also  outside  the 
metal  field,  and  constitutes  in  each  case  a  distinct  innovation  as  well  as  a  distinct  industrial 
problem.     The  prices  of  scrap  are  more  sensitive  to  the  course  of  cyclical  phases  than  any 
other  commodity  prices  and,  as  has  been  pointed  out  by  the  Berlin  Institute  (e.g.,  in  1926), 
the  relation  between  scrap-  and  pig-iron  prices  is  a  good  index,  even  forecaster,  of  business 
situations. 

2  Steel  prices  never  came  down  to  the  1914  average  and  kept  substantially  above  it 
even  during  the  crisis.     Their  behavior  is  interesting  because  of  the  differences  in  the  price 
policy  of  different  concerns,  which  make  it  difficult  to  speak  of  one  price  at  a  time.    The 
lead  of  the  United  States  Steel  Corporation  was  not  consistently  followed  by  the  inde- 
pendents.    But  neither  did  the  latter  do  what  one  might  expect  from  younger  and  more 
vital  concerns,  viz.,  take  an  initiative  in  reducing  prices.     On  several  occasions  they  were 
more  eager  to  advance  and  more  reluctant  to  reduce  them  than  the  United  States  Steel 
Corporation  was. 


782  BUSINESS  CYCLES 

through  1934. 1  Thus  it  failed  to  fall  in  1930,  when  it  would  have  done 
so  under  competitive  conditions.  But  the  profits  made  are  not  in  them- 
selves sufficient  to  prove  that  in  the  long  run  prices  were,  given  the  pro- 
tection, substantially  above — they  even  may  have  been  below — that 
level  at  which  they  would  have  moved  had  competitive  conditions  pre- 
vailed from  the  outset,  provided  we  include  in  the  latter  the  degree  of 
productive  efficiency  compatible  with  the  competitive  scale  of  individual 
firms.  Nor  does  it  follow  that,  if  all  or  most  industries  had  been  organized 
in  the  same  manner,  they  would  have  still  found  it  to  their  interest  to 
adopt  the  same  policy  of  price  stability.2 

The  war  for  obvious  reasons  brought  a  large  expansion  in  the  con- 
sumption of  copper,  to  which  the  new  mines  and  mining  methods  (see 
Chap.  VII)  were,  however,  fully  equal,  so  much  so  that  already  in  1917 
efforts  were  made  to  fix  prices.  They  were  followed  by  others  throughout 
the  twenties,  which  in  fact  succeeded  in  keeping  price  fairly  stable  at 
about  prewar  level  (12  to  15  cents)  from  1921  to  1928.  At  the  same  time 
costs  were  being  incessantly  reduced  by  further  development  of  large- 
scale  mining  methods  and  by  new  processes  of  smelting  and  refining  as 
well  as  by  the  discovery  of  better  deposits.  There  were  also  important 
horizontal  combinations — among  mining  companies,  smelters  and  refiners, 
and  copper-  and  brass-producing  concerns — and  vertical  ones.  Thus 
an  untenable  situation  developed — indicated  by  the  fact  that  production 
kept  persistently  above  consumption,  stocks  being  well  above  prewar 
level  throughout — under  what  for  a  decade  seemed  a  prosperous  surface : 
one  of  the  weak  spots  that  were  to  contribute  their  share  to  the  processes 
of  the  world  crisis.  Secondary  copper — in  1929,  production  of  secondary 
copper  from  scrap  was  40  per  cent  of  smelter's  production  of  new  copper 
from  domestic  ore — and  the  output  of  low-cost  mines  in  Canada,  Katanga, 
and  Northern  Rhodesia  did  the  rest.  The  formation  of  an  international 
export  cartel  under  the  Webb-Pomerene  Act  (Copper  Exporters,  Inc., 
October  1926,  which  "controlled"  90  per  cent  of  the  world's  production) 
only  deferred  the  catastrophe. 

We  have  here  an  extremely  interesting  case  of  an  otherwise  perfectly 
unfettered  process  of  capitalist  innovation,  which  was  interfered  with 
only  by  capitalist  interests  themselves  and  by  these  again  only  through 
an  attempt  to  put  out  of  operation  a  single  element  of  the  mechanism, 

1  Data  are  from  the  Engineering  and  Mining  Journal.     The  American  Metal  Market 
gives  for  the  same  year  an  average  of  21.6. 

2  In  that  case  reduction  by  agreement  of  all  those  prices  would  have  been  practicable 
and  might  have  suggested  itself  because,  with  respect  to  all  of  them,  demand  would  not 
have  been  as  inelastic  as  it  was  with  respect  to  each  of  them  taken  singly.    In  the  case  of 
public  price  fixing,  on  the  other  hand,  that  stability  might  have  become  an  end  in  itself, 
as  it  did  in  most  cases  of  price  regulation  by  public  authority. 


1919-1929  783 

viz.,  the  effect  of  "progress"  on  price.  It  is  worth  while  to  consider 
what  the  course  of  events  might  have  been  without  such  an  attempt. 
Prices  would  certainly  have  fallen  and  it  may  well  be  that  especially  in 
the  short  run  this  would  not  have  increased  consumption  appreciably. 
But  this  is  not  the  point.  Mines  and  refineries  which  actually  were 
kept  alive  would  have  been  eliminated  in  any  case,  though  of  course 
more  of  them  if  demand  was  really  inelastic  in  the  relevant  range  than 
if  it  was  not.  And  this  would  have  first  eliminated  waste — for  it  is 
social  waste  to  work  a  mine  or  refinery  which  can  be  worked  only  at  an 
"artificial  price" — and  second  helped  to  tone  down  the  prosperities  of 
the  twenties,  to  spread  the  work  of  readjustment  and  pro  tanto  to  mitigate 
the  subsequent  crisis.  If  all  this  be  impossible  or  more  than  the  system 
can  stand,  public  regulation  or  ownership  is,  under  such  conditions — 
the  reader  should  carefully  analyze  in  what  they  differ  from  those  of  the 
aluminum  case — in  fact  the  only  alternative  to  violent  breakdowns, 
though  not  necessarily  a  remedy  for  this  type  of  economic  waste. 

e.  While  it  is  believed  that  the  above  exposition  brings  out  the  salient 
features  of  the  industrial  processes  of  the  decade  under  discussion,  estab- 
lishes what  we  have  set  out  to  establish,  and  suffices  for  our  purposes — 
particularly  for  the  purpose  of  interpreting  the  business  situations  of 
the  period  and  the  behavior  of  aggregative  time  series — it  must  once 
more  be  emphasized  how  very  incomplete  it  is.  It  even  leaves  out  some 
major  elements — natural-gas-pipe  lines  (1927)  have,  for  instance,  not 
even  been  mentioned — and  practically  all  those  minor  ones1  the  sum 
total  of  which  is  particularly  important  in  the  downgrade  of  a  Kondratieff . 
Knowledge  of  the  full  extent  of  the  revolution  which  that  period  wit- 
nessed both  in  the  methods  of  production  and  commerce  and  in  the 
structure  of  the  budgets  of  households,  and  an  adequate  analysis,  in  the 
light  of  it,  both  of  the  period  itself  and  of  its  aftermath,  would  presuppose 
very  many  case  studies  beyond  those  we  at  present  have.2  Nevertheless, 
the  main  features  stand  out  unmistakably  and  can  be  further  illustrated 
by  a  few  facts  from  the  Abstract  of  the  Census  of  1930. 3  This  census 

1  A  major  movement,  which  however  resolves  itself  into  an  almost  infinite  number  of 
small  ones,  is  what  may  be  called  Taylorization.     Its  spread  during  our  period  is  a  typical 
consequence  of  the  struggle  for  survival  amidst  the  readjustments  of  downgrades.     The 
pressure  of  this  country's  wage  level  adding  momentum,  this  type  of  rationalization  of 
every  job  was  in  many  cases  more  effective  in  reducing  costs  per  unit  of  product  than 
fundamental  innovations  could  have  been — and  in  all  cases  highly  significant  from  our 
standpoint.     It  is  a  special  case  of  a  class  of  which  the  efforts  to  utilize  scrap  and  waste 
are  another. 

2  The  most  important  achievement  in  that  field  is  Mr.  H.  Jerome's  Mechanization  in 
Industry,  herewith  recommended  once  more  to  the  attention  of  the  reader  as  a  storehouse 
of  examples  of  typical  downgrade  improvements  by  which  to  test  our  interpretation  of  the 
period. 

»  Pp.  744-759. 


784  BUSINESS  CYCLES 

lists  103  industries  each  of  which  had  in  1929  a  Value  Added  of  over 
50  million  dollars  and  also  was  independently  listed  in  1919.  On  the 
average  (unweighted)  Value  Added  increased  between  these  two  years 
by  29  and  the  ratio  between  Value  Added  and  Pay  Roll  by  16  per  cent.1 
First,  however,  we  are  interested  in  those  industries  which,  while  pro- 
ducing a  Value  Added  of  over  50  million  dollars  in  1929,  do  not  inde- 
pendently occur  in  1919,  since  this  in  itself  proves  a  very  rapid  rate  of 
growth.  There  are  16  of  them:  beverages,2  food  preparations,  millinery, 
motion  pictures  (excluding  projection  in  theaters),  paper,  pulp,  rayon, 
refrigerators,  rubber  tires  and  inner  tubes,  other  rubber  goods  (excluding 
boots),  and  typewriters  being  the  most  significant  ones.  Besides  repeat- 
ing cases  which  we  know  already,  this  list  adds  a  few  new  elements  to 
our  picture.  Of  particular  significance  is  the  suggestion,  which  under- 
lines one  feature  of  that  phase  of  the  long  cycle,  of  industries  which 
expanded  simply  in  response  to  the  increase  of  consumers'  real  purchasing 
power  and  without  any  particularly  strong  impulse  of  their  own. 

This  suggestion  grows  stronger  still  if,  second,  we  glance  at  those 
among  the  103  industries  that  display  an  increase  in  Value  Added  of, 
say,  more  than  100  per  cent.  Besides  aircraft  and  parts,  which  heads 
the  list  (510  per  cent),  electrical  machinery  and  supplies,  aluminum 
manufacture,  motor  vehicle  bodies  and  parts,  which  we  would  expect 
to  find,  we  also  meet  perfumes  and  cosmetics,  signs  and  advertising 
novelties,  concrete  products,  flavoring  extracts  and  sirups,  photo-en- 
graving not  done  in  printing  establishments,  house-furnishing  goods, 
ice  cream,  printing  and  publishing  (newspaper  and  periodical).  Patent 
medicines,  soap,  cigars  and  cigarettes,  cereal  preparations,  bakery  prod- 
ucts, while  not  reaching  the  100  per  cent  line,  yet  increased  their  Value 
Added  by  much  more  than  the  average  figure  so  as  to  reinforce  the  evi- 
dence. Third,  we  will  note  some  of  those  industries  the  Value  Added 
of  which  decreased  by  more  than  10  per  cent:  shipbuilding,  locomotives 
(not  made  in  railroad  repair  shops),  railway  cars,  pianos,  phonographs, 
leather,  beet  sugar  (Value  Added  of  the  cane-sugar  industry  remained 
at  the  1919  figure),  cotton,  woolen  and  worsted  goods,  flour. 

rThe  limitations  of  Value  Added  as  a  measure  of  the  development  of  an  industry, 
and  of  the  ratio  between  Value  Added  and  Pay  Roll  as  a  measure  of  innovation  or  induced 
and  completing  rationalization  are  too  obvious  to  be  explicitly  stated.  It  is  easy  to  see, 
however,  in  what  sense  both  are  relevant  to  our  subject. 

2  The  industry  of  nonalcoholic  beverages  was,  of  course,  conditioned  by  prohibition. 
There  were,  however,  several  minor  innovations,  some  of  them  of  a  purely  commercial 
type.  But  prohibition  also  conditioned  the  creation  of  organizations  for  the  illicit  trade 
in  alcoholic  beverages,  which  constituted  innovations  and  a  new  industry  in  our  sense. 
The  somewhat  unconventional  character  of  this  industry  only  serves  to  illustrate  some 
aspects  of  our  concept  of  entrepreneurship.  If  we  exclude  it,  this  is  only  because  the  writer 
had  to  come  to  the  conclusion  that  data  were  too  unreliable  for  use. 


1919-1929  785 

We  cannot  expect,  nor  do  we  find,  significant  correlation  between  per 
cent  increase  in  Value  Added  and  per  cent  increase  in  the  ratio  of  Value 
Added  to  Pay  Roll.  Also,  some  new  or  relatively  new  industries,  such 
as  motor  bodies  or  aluminum  manufactures,  understandably  show  very 
little  signs  of  the  effects  of  labor-saving  devices  on  that  ratio,  although, 
of  course,  others,  such  as  motorcars  or  aircraft,  rank  high,  and  some  of 
the  old  and  conspicuously  noninnovating  ones,  such  as  house-furnishing 
goods  (1  per  cent  decrease),  rank  low  in  this  respect.  It  is,  however, 
instructive  to  observe — and  tells  a  great  deal  about  the  general  character 
of  the  industrial  processes  of  that  time — how  much  labor-saving  rationali- 
zation went  on  outside  of  the  great  lines  of  innovation.  Thus  the  ratio 
between  Value  Added  and  Pay  Roll  increased  by  120  per  cent  with 
cigars  and  cigarettes,  85  per  cent  with  soap  and  with  coke  (excluding 
gashouse  coke,)  71  per  cent  with  cereal  preparations,  61  per  cent  with 
manufactured  gas,  52  per  cent  with  cutlery  and  edge  tools,  52  per  cent 
with  ice  cream.  Even  for  tin  cans  that  figure  is  still  33;  for  patent 
medicines,  druggists'  preparations,  coffee  roasting  and  grinding  32; 
cane-sugar  refining  28;  meat  packing  (wholesale)  26;  butter  26;  cement 
and  concrete  products  22;  perfumes  and  women's  clothing  17.  Only  in  a 
minority  of  cases — printing  would  be  one — was  this  due  to  substantive 
novelties  that  we  have  simply  been  unable  to  mention.  In  the  main  it 
was  the  result  of  that  systematic  effort  to  fight,  under  the  pressure  of 
a  price  level  that  tended  to  fall,  each  cost  item  by  exploring  every  detail 
of  the  productive  and  the  commercial  process  and  by  applying  and  devel- 
oping techniques  the  fundamentals  of  which  were  fully  established  before 
the  war,  but  which  in  many  cases  involved  not  only  technological  improve- 
ment in  existing  plant  but  also  the  erection  of  new,  highly  mechanized 
plant.1 

/.  As  pointed  out  elsewhere,  it  would  for  this  country  be  possible  to 
carry  our  count  of  Juglars  and  Kitchins  through  the  war,  which  never 
succeeded  in  blotting  them  out  completely.  But  in  order  to  avoid  a 
statement  which  cannot  be  usefully  made  unless  it  be  more  fully  estab- 
lished than  is  possible  here,  we  will  now  not  go  beyond  saying  that  the 
end  of  1916  and  the  beginning  of  1917  might,  but  for  the  war,  have  wit- 
nessed what  according  to  our  terminology  would  have  been  the  beginning 
of  the  prosperity  phase  of  the  third  Juglar  of  this  Kondratieff,  and  that 
in  this  case  the  crisis  of  1921  would  have  occurred  exactly  when  our 
schema  would  have  led  us  to  expect  it,  i.e.,  when  that  Juglar  turned  from 

1See  E.  F.  Baker,  Unemployment  and  Technical  Progress  in  Commercial  Printing, 
American  Economic  Review,  September  1930,  and  Technological  Change  and  Organized 
Labor  in  Commercial  Printing,  American  Economic  Review,  December  1932.  The  great 
increase  in  our  ratio  under  the  heading  Cigars  and  Cigarettes  was  in  part  due  to  the  new 
machine  introduced  into  cigar  making. 


786  BUSINESS  CYCLES 

recession  into  depression.  Even  if  we  wished  to  press  this — which  we 
do  not — it  would  leave  us  all  the  freedom  in  the  world  to  take  into  account 
the  effects  of  war  demand  and  war  finance,  of  the  shock  imparted  to  the 
war  structure  by  the  armistice — i.e.,  the  four  months  of  dullness  or 
wavering  rather  than  collapse  which  followed  upon  it — of  foreign  and 
domestic  postwar  demand  producing  the  boom  of  1919  and,  finally,  of 
all  that  partial  liquidation  of  both  war  and  postwar  situations  contributed 
to  the  slump  of  1920-1921.  *  To  what  has  on  various  occasions  already 
been  said  about  the  latter  it  is  necessary  to  add  but  two  comments. 

First,  however  clear  it  is  that  that  slump  was  primarily  a  process  of 
liquidation  of  war  effects  and  a  reaction  to  the  boom  of  1919 — which  in 
turn  had  little  if  anything  to  do  with  innovation — yet  the  presence,  and 
businessmen's  awareness  of  the  presence,  of  a  new  industrial  situation — 
a  situation  which  was  new  in  the  sense  that  the  consequences  of  prewar 
innovations  had  profoundly  altered  the  cost  structure — had  much  to  do 
with  the  severe  restriction  in  output  of  manufactures  which  first  began  in 
January  and  again,  after  a  rebound,  in  March  1920,  in  the  face  of  the 
facts  that  retail  sales  had  throughout  1919  increased  at  a  greater  rate 
than  had  production,2  that  the  export  trade  as  yet  showed  no  signs  of 
slackening,  that  prices  continued  to  rise.  Banks,  moreover,  were,  by 
the  influx  of  gold  (gold  imports  in  1921  amounted  to  nearly  750  millions) 
and  by  the  reduction  of  government  debt  by  about  1.2  billion  dollars 
between  June  30,  1919  and  June  30,  1920,  enabled  to  increase  their  loans 
by  about  1.5  billion  dollars  during  the  same  time.3  All  this  puts  some 
of  the  most  popular  theories  out  of  court  in  this  case.  We  have  once 
more  an  instance  of  "business  deflating  itself"  without  any  serious 
outside  pressure,  and  we  see  again  that  this  could  have  been  prevented 
only  by  continuing  government  expenditure  at  the  war  level  or  a  level 
still  higher.  The  question  why  business  deflated  itself  cannot  be  given, 
even  in  this  case,  without  reference  to  our  mechanism. 

Second,  the  reaction  was  sharp  and  unimpeded  and,  because  it  was 
sharp  and  unimpeded,  relatively  short.  Prices  and  wages  were  allowed 
to  drop  drastically,  liquidation  of  commodity  stocks  and  debts  proceeded 

1See  W.  M.  Persons,  The  Crisis  of  1920  in  the  United  States,  American  Economic 
Association,  84th  Annual  Meeting  (December  1921).  The  picture  which  that  paper  draws 
of  the  situation  at  the  time  is  perfectly  sufficient  for  our  purpose,  so  that  this  reference  may 
relieve  our  text  of  the  necessity  of  going  into  details. 

2  This  has  been  emphasized  by  Professor  Slichter,  The  Period  1919-1936  in  the  United 
States,  Review  of  Economic  Statistics,  February  1937.     Although  considerations  of  space 
forbid  the  presentation  of  the  year-to-year  analysis  of  business  situations  from  1921  to  1929, 
carried  out  for  the  purposes  of  this  book  by  Professor  E.  M.  Hoover,  comparison  between 
Professor  Slichter's  business  history  of  the  period  and  the  remarks  that  follow  in  our  text  is, 
nevertheless,  invited. 

3  Federal  reserve  rediscounts  reached  a  peak  of  2,827  million  dollars  on  Nov.  5,  1920. 


1919-1929  787 

rapidly,  elimination  of  firms — over  8  per  cent  of  the  manufacturing  firms 
which  were  in  business  in  1919  had  disappeared  by  1921 — was  prompt, 
money  rates  fell,  credit  was  readily  available,  and  the  situation  began  to 
stabilize  itself  in  April  1921,  the  textile  and  clothing  industries,  which 
had  expanded  first  in  1919  and  fallen  first  in  1920,  being  among  the  first 
to  revive.1  The  resulting  price  relations  differed  greatly  from  those  of 
1913  and  struck  many  observers  as  entirely  abnormal.  But  the  change 
was  largely,  though  not  wholly,  one  of  adaptation  to  new  conditions. 

Though  improvement  slackened  in  October  and  many  signs  of  con- 
tinuing liquidation — e.g.y  further  reduction  in  wages — outlasted  the  sum- 
mer of  1922,  "deep  depression"  was  over  by  December  1921.  In  April 
1922  the  automobile  and  tire  industries  experienced  shortage  of  labor, 
while  stock  issues  had  already  revived  in  January:  it  was  then  that  the 
boom  in  public  utility  stocks  began.  Prices  of  steel,  tires,  glass,  and 
oil  rose  in  the  fall,  while  those  of  gasoline,  automobiles — the  price  of 
tractors  had  been  reduced  before  by  action  of  the  Ford  concern — cement, 
and  foods  fell.  In  December  1922  the  oil  industry  was  breaking  all 
previous  records  in  output.  The  fact  that  such  should  have  been  the 
situation  only  one  year  after  what  nobody  doubts  had  been  a  major 
crisis  and  in  the  presence  of  many  depressive  symptoms — federal  reserve 
rediscounts,  however,  which  fell  to  a  trough  of  380  million  dollars  can- 
not, owing  to  the  gold  influx,  be  counted  among  them — is  full  of  potential 
lessons  which  are  as  obvious  as  they  are  useless.  The  case  also  shows 
better  than  any  theory  could  how  the  system  pulls  out  of  troughs  under 
its  own  steam  and  how  it  succeeds  in  doing  so  while  price  level  is  still 
falling. 

Our  diagnosis  then,  which  it  would  take  very  strong  preconceptions 
to  doubt,  is  simple:  abnormally  short  depression  phase  of  the  Juglar, 
lasting  from  the  fall  of  1920  to  July  of  1922,  owing  to  abnormally  effec- 
tive liquidation.  Alternatively,  we  could  express  the  same  facts  by 
saying  that  the  depression  phase  lasted  to  the  end  of  1922  but  that  its 
work  had  so  effectively  been  done  by  May  1921  that,  the  ground  being 
cleared,  the  prosperity  phase  of  the  third  Kitchin,  which,  as  we  know,  still 
belongs  to  the  depression  phase  of  its  Juglar,  had  unusual  opportunities 
of  asserting  itself.  In  any  case  this  Kitchin  stands  out  unmistakably, 
and  there  is  no  reason  why  we  should  not  so  call  an  undoubtedly  short 

1  Of  course,  the  above  statement  excludes  those  commodities  which,  like  oil,  electricity, 
gas,  meat,  motors,  had  either  not  fallen  at  all  or  fallen  only  a  little.  The  activity  in  the 
petroleum  industry  has  been  emphasized  before.  Several  automobile  factories  which  had 
been  shut  down  reopened  in  April.  The  tin  can  and  cigar  industries  (new  plant  of  the 
Continental  Can,  activity  of  the  Bayuk  Cigar  Company)  were  also  active.  So  were  the 
shoe  and  leather  and  the  drug  industries.  One  cigarette  company  reported  that  in  the  first 
months  of  1921  it  had  earned  more  than  double  the  whole  year's  dividend  requirements. 


788  BUSINESS  CYCLES 

cycle  which  is  universally  recognized — even  if  differently  dated  by  stu- 
dents who  count  from  trough  to  trough — and  which  completely  answers 
expectation  from  our  schema  both  as  to  formal  characteristics — length 
and  location  included — and  as  to  industrial  meaning.  On  the  other  two 
Kitchins  which  within  that  Juglar  ought  to  have  preceded  this  one,  we 
will  not  insist,  although  it  would  not  be  difficult  to  establish  them 
statistically. 

What  followed,  from  either  the  middle  or  the  end  of  1922  on,  either 
is  or  very  much  looks  like  a  normal  Juglar  recovery,  which  lasted  to  the 
autumn  of  1925.  Our  schema  would  lead  us  to  expect  that  it  contained 
a  setback  owing  to  the  Kitchin  depression  which  it  tells  us  should  have 
occurred.  It  did  occur.  After  it  had  run  its  course,  recovery  resumed 
and  from  August  1925  on  both  Kitchin  and  Juglar  were  shading  off  into 
the  prosperity  phases  of  what  then  would  be  the  fourth  Juglar  and  its 
first  Kitchin.1  A  few  additional  facts  may  be  useful.  In  the  first 
quarter  of  1923  the  upswing  reached  its  peak.  Unemployment  was  at 
low  ebb — some  people  spoke  of  its  being  "absent" — in  February.  Most 
prices,  especially  those  of  metals  and  building  materials,  tended  upward; 
a  record  year  for  construction  was  correctly  foreseen.  Steel — 19  new 
steel  furnaces  were  built  in  1923 — coal,  and  cotton  textiles  expanded. 
Four  new  power  stations  were  announced  for  construction.  Railroads 
"came  into  their  own  again"  and  gave  orders.  Everything  except  agri- 
culture and  ship-building  boomed.  Wage  rates  rose  strongly.  April 
saw  record  figures  but  also  a  break  in  the  stock  market.  Signs  of  slack- 
ening activity  began  to  show  by  August,  attributed  as  usual  to  external 
factors,  and  by  December  expectations  were  not  very  optimistic.  They 
were  borne  out  by  the  state  of  things  during  the  spring  and  summer  of 
1924.  More  important  than  the  uncertainties  incident  to  the  presidential 
campaign  were  the — understandable — reactions  in  the  automobile  and 
oil  industries.  Steel  consumption,  railroad  traffic,  employment,  and 
prices  fell.2  Nothing  very  serious  happened,  however;  failures  of  com- 
mercial firms  were  but  insignificantly  above  the  1923  figure;  residential 
building  kept  up  well,  and  so  did  power  production,  the  radio  industry, 
and  other  lines.  The  first  two  months  of  1925  were  disappointing — with 
employment  in  most  industries  below  what  it  had  been  a  year  before — 
and  a  collapse  of  the  stock  market  followed  in  March.  During  the  second 
quarter  business  was  described  as  steady  but  "spotty."  New  financing 
and  other  indications  of  imminent  prosperity  asserted  themselves  under 
this  surface,  however,  and,  with  the  help  of  improvement  in  the  agrarian 

1  Professor  Hoover  wrote  in  the  report  already  mentioned:  "By  August  the  upsurge  was 
unmistakable  and  the  word  prosperity  began  to  creep  modestly  back  into  the  vocabulary 
of  the  [Financial  and  Commercial}  Chronicle" 

2  There  was  an  abortive  rally  in  November. 


1919-1929  789 

sphere  and  the — largely  speculative — land  boom  ("real  estate  subdi- 
vided"), the  fall  wore  on  amidst  record  investment,  bank-clearing 
(October)  and  construction  figures,  rising  money  rates  and  steel  prices, 
stock  market  excesses  (October),  failures  at  record  low.1 

The  explanation  of  all  this — the  "ignition" — will  be  found  on  referring 
to  the  above  survey  of  the  fundamental  industrial  processes  of  the  period. 
They  clearly  change  during  those  months  owing  to  the  influence  of  several 
new  impulses — while  others,  such  as  residential  bulding,  lost  force — and, 
by  conforming  exactly  to  what  we  mean  by  a  Juglar  prosperity,  justify 
our  dating.  So  much  is  provable  and  indeed  obvious.  But  we  will 
for  a  moment  trust  our  schema  to  the  point  of  absurdity  and  try  to 
"predict"  the  subsequent  business  situations  on  the  assumptions,  first, 
that  the  fourth  Juglar  started  with  the  fourth  quarter  of  1925 — although 
we  know  that  no  such  exactness  is  possible  in  historical  analysis — second, 
that  its  duration  was  to  be  exactly  9^  years — roughly  the  average 
duration  of  prewar  Juglar s — and  the  duration  of  its  Kitchins  exactly 
38  months;  third,  that  all  the  Juglar  and  Kitchin  phases  were  of  exactly 
equal  length.  This  absurd  experiment  yields  the  following  results:  the 
Juglar  prosperity  lasting  into  February  1928  should  be  interrupted  by  a 
Kitchin  depression  from  May  1927  to  the  middle  of  February  1928; 
and  the  recovery  of  this  Kitchin — to  the  end  of  November  1928 — and  the 
prosperity — to  the  middle  of  September  1929 — and  the  recession  of  the 
second  Kitchin  should  then  run  their  course  within  the  Juglar  recession 
ending  with  June  1930.  At  that  date  both  the  Juglar  and  the  Kitchin 
should  enter  upon  their  depression  phases  on  a  Kondratieff  that  had 
already  entered  upon  its  own,2  so  that  the  configuration  of  1873  would  be 
repeated.  The  reader  will  realize  that  no  value  attaches  to,  and  no 
significance  is  claimed  for,  the  exact  dates.  But  he  will  also  realize  that 
the  absurdity  stops  at  the  assumptions  which  are  responsible  for  the 
exact  dates.  Stripped  of  this  unwarranted  exactitude  and  confined  to 
essentials,  the  "predictions"  of  the  schema  are  not  absurd  but  on  the 
contrary  tell  several  important  truths — and  not  a  single  untruth.  It 

1  Mr.  Thorp's  Annals  closing  with  1925,  we  may  compare  his  description  with  ours. 
Bearing  in  mind  differences  in  terminology,  we  find  almost  complete  agreement.     He  notes 
what  above  was  described  as  dullness  in  1918  ("recession")  and  the  boom  of  1919.     1920  is 
by  him  described  by  prosperity-recession-depression,  1921  by  depression.     1922  is  a  year 
first  of  revival  then  of  prosperity,  which  is  obviously  meant  to  convey  exactly  what  was 
meant  above.     In  1923  there  was  prosperity-recession  and  in  1924  mild  depression-revival. 
1925  Mr.  Thorp  describes,  as  we  should  expect  he  would,  as  a  year  of  prosperity  without 
qualification.     The  present  writer  is  not  sure  whether  employment  was  really  "full"  or 
that  the  rise  in  bond  prices  accords  with  the  pattern  of  prosperity.     But  there  is  no  material 
difference  in  diagnosis. 

2  It  will  be  remembered  that  according  to  our  schema  Kondratieff  depressions  begin 
with  Juglar  prosperities.     The  depression  phase  of  the  current  Kondratieff  would,  hence, 
date  from  the  fall  of  1925. 


790  BUSINESS  CYCLES 

should  be  added  that  the  comparative  severity  of  the  setback  in  1927, 
which  was  to  occur  and  did  in  fact  occur  within  the  prosperity  phase 
of  the  Juglar,  does  not  in  itself  run  counter  to  expectation :  the  depression 
of  a  Kitchin,  located  as  that  one  was,  after  the  end  of  the  Kondratieff 
recession,  should  be  well  marked.1  On  the  other  hand,  the  boom  of 
1928-1929  was  more  violent  than  our  schema  leads  us  to  expect,  which 
in  explanation  has  but  a  Kitchin  recovery  and  a  Kitchin  prosperity  to 
offer.  This  may,  however,  be  accounted  for  by  certain  autonomous 
monetary  factors  and  the  influence  of  the  speculative  mania,  of  which  the 
first  do  not  form  part  of  our  model  and  the  second — also  present  in  1872- 
1873 — is  always  an  irregular  factor. 

The  stock  market  suffered  collapse  in  February  1926.  But  this  is 
merely  a  normal  incident  of  a  Juglar  prosperity  outgrowing  its  initial 
stage.  A  no  less  regular  phenomenon  was,  on  a  Kondratieff  depression, 
the  tendency  of  prices  to  sag.  If  business  conditions  began  to  display 
signs  of  "relaxation"  already  by  April,  when  automobile  concerns  did  not 
do  so  well  as  they  had  a  year  before  and  the  cotton,  silk,  sugar,  and  other 
industries  headed  toward  curtailment,  and  if  in  May  there  was  also  a 
decline  in  steel  production,  this  is  sufficiently  accounted  for  by  previous 
speculative  excesses,  in  particular  by  the  passing  of  the  real  estate  boom. 
The  stock  exchange  recovered  by  June,  and  almost  everything  was  again 
at  prosperity  levels  by  August,  motorcars  and  textiles  included,  furniture 
enjoying  record  profits.  Oil  developments  in  California,  in  North  Texas, 
and  on  the  Gulf  Coast  did  not  entail  any  large  increase  in  stocks.  By 
October,  however,  the  1927  setback  came  definitely  in  sight.  The  stock 
market  discounted  it,  bank  debits  were  running  from  5  to  9  per  cent 
below  the  figures  of  the  preceding  year,  demand  for  steel  dropped  until 
operations  were  at  65  per  cent  capacity.  Failures  increased.  Car  load- 
ings also  were  at  the  end  of  what  nevertheless  was  a  record  year  at  a 
lower  figure  than  they  had  been  in  1925.  Some  anxiety  was  felt  about 
installment  sales.2  The  agrarian  situation  had  also  become  more 
unfavorable. 

1  Professor  Mitchell  dates  one  of  his  cycles  from  1927.     Since  the  writer  naturally 
wishes  to  differ  as  little  as  possible  from  so  outstanding  an  authority,  he  begs  to  emphasize 
that  no  difference  in  diagnosis  of  situations  is  implied  in  such  dating,  because  it  is  simply 
the  consequence  of  Professor  Mitchell's  principles  to  count  from  trough  to  trough  and  to 
recognize  but  a  single  type  of  cycles.     The  particular  pattern  under  discussion  seems  well 
qualified  for  suggesting  that  some  important  elements  of  reality  are  being  missed  if  we  put 
troughs  such  as  occurred  in  1924  and  1927  on  the  same  level  with  troughs  such  as  occurred, 
say,  in  1908  and  1921,  and  these  again  on  the  same  level  with  the  troughs  of  1875  and  1932, 
and  that  the  distinction  of  cycles  of  different  type  seems  the  natural  way  of  recognizing 
these  very  real  differences. 

2  Consumption  was  at  a  high  level.     But  no  less  than  10  per  cent  of  the  cars  sold 
during  that  year  could  not  be  paid  for  and  had  to  be  repossessed  by  the  dealers — a  good 
illustration  of  the  theory  of  underspending  or  oversaving. 


1919-1929  791 

Until  May  1927,  however,  general  business  kept  on  a  high  level  in 
spite  of  all  that  and  even  improved,  several  new  things — the  Chevrolet 
and  Frigidaire  successes,  motion  pictures,  the  North  Carolina  power 
plant,  a  number  of  smaller  events — supplying  impulse.  In  April  busi- 
ness was  prosperous.  But  then  a  definite  decline  set  in — which  we  may 
identify  as  a  Kitchin  depression — intensified  by  widespread  recognition 
of  unsound  practice  in  many  fields,  with  retail  and  wholesale  trade  at  a 
lower  level,  many  failures,  and  cautious  reserve  in  large-scale  business. 
Building,  the  condition  of  which  was  complicated  by  the  liquidation  of  the 
Florida  boom,  was  a  particularly  weak  spot.  The  fall  in  automobile 
production  was,  of  course,  due  to  the  reconstruction  of  the  Ford  plant.1 
The  Mississippi  flood,  while  it  also  explains  some  things  about  the  behavior 
of  physical  indices,  has  in  other  respects  to  be  listed  as  an  impetus.  There 
was  no  slump.  Good  business  in  the  cotton,  rayon,  and  shoe  industries 
and  a  continuing  stream  of  new  things — Dicsel-engined  locomotives, 
gas-pipe  lines,  the  refrigerator  merger,  development  of  the  Kraft  paper 
industry  in  the  South,  radios — were  features  throughout.  By  Decem- 
ber improvement  was  almost  general,  although  employment  in  building 
was  still  12  per  cent  below  that  of  a  year  before. 

Railroad  earnings,  steel  production,  and  gasoline  markets  improved 
in  January  1928,  and  the  "bankers'  loans  boom"  in  the  stock  market 
was  getting  under  sail  with  automobile,  copper,  and  rubber  stocks  leading. 
But  the  general  business  situation2  behaved  until  March  (even  in  March 
there  was  considerable  unemployment)  in  a  manner  which  is  in  our 
terminology  not  badly  rendered  by  the  phrase  "conditions  of  Kitchin 
depression  relieved  by  a  Juglar  prosperity."  In  April,  however,  steel 
was  at  record  rates.  So  was  tire  production  and  by  June  everything — 
building,  the  automobile  (contraseasonally)  and  the  oil  industry  included. 

1  But  the  writer  fails  to  understand  how  some  observers  could  have  attributed  the  1927 
depression  wholly  to  that  fact. 

2  The  contemporaneous  comments  of  the  public  press,  and  business  opinion  as  revealed 
by  public  utterances  of  executives  and  so  on,  quite  correctly  realized  and  sized  up  some  of 
the  fundamental  elements  of  that  situation.     The  writer's  impression  is  that  there  was  not 
only  more  insight  but  also  more  foresight  in  considerable  sectors  of  the  business  community 
than  is  in  general  realized,  and  also  more  than  is  compatible  with  some  theories  which 
overstress  the  element  of  error  in  the  explanation  of  depressions.     But  one  also  finds  the 
tendency,  especially  in  press  comments  the  authors  of  which  aim  at  semiscientific  explana- 
tions, to  overstress  external  and  banking  factors.     The  preceding  note  affords  an  example 
of  the  former  type.     The  latter  type  is  instanced  by  the  emphasis  on  Federal  Reserve 
Board  policy  in  explanation  of  the  state  of  things  in  the  first  two  months  of  1925  and  again 
the  first  two  months  of  1928.     In  both  cases  the  industrial  process  itself  supplies  much 
more  plausible  explanations.     But  it  is  easy  to  understand  how  tempting  that  explanation 
was,  especially  for  writers  whose  contact  with  business  was  primarily  through  the  money 
market  and  who  were  strongly  influenced  by  the  scientific  and  pseudoscientific  "theories" 
of  the  epoch. 


792  BUSINESS  CYCLES 

Symptoms  of  "high"  prosperity  then  went  on  intensifying  themselves 
until  October,  when  mail-order  sales  broke  all  records.  Construction  of 
new  plant — for  old  as  well  as  for  some  new  purposes — new  financing — 
taking  advantage  of  the  stock-market  boom — dividends,  money  rates, 
and  so  on  were  all  in  keeping  with  the  rest  of  the  picture,  into  the  details 
of  which  we  are  unable  to  go.  There  were  two  apparently  discordant 
elements,  however.  First,  unemployment  increased.  Second,  com- 
modity markets  though  buoyant  were  not  really  sellers'  markets:  the 
almost  desperate  efforts  made  by  the  sales  organizations  of  big  and  small 
concerns  and  the  fact  that  such  increase  in  prices  as  occurred  was  insig- 
nificant, while  many  important  prices  had  to  be  reduced,  indicate  a 
certain  strain  in  the  system. 

Now  if  the  reader  will  remember  the  writer's  various  attempts  to 
convey  his  idea  of  a  Juglar  recession,  he  will  appreciate  the  warrant  for 
expressing  that  state  of  things  by  saying  that  it  was  exactly  what  we 
should  expect  from  a  Juglar  recession  on  a  Kondratieff  depression  coupled 
with  the  two  positive  phases — strictly,  according  to  our  experimental 
schema,  the  recovery  phase  only — of  a  unit  of  the  short  cycle:  good  and 
expanding  business  accompanied  by  increasing  unemployment  and  by 
that  strain  which  is  the  consequence  of  the  "avalanche  of  goods"  smash- 
ing its  way  through  the  resisting  framework  of  the  existing  industrial 
structure — this  is  precisely  the  picture  which  would  result  from  that 
particular  juncture.  As  stated  before,  however,  there  is  no  doubt  that 
the  developments  between  April  1928  and  August  1929  added  to  the 
situation  many  of  the  untenable  elements1  which  subsequently  served 
to  intensify  the  crisis. 

October  1928  brought  the  first  symptom  of  slackening  activity,  which 
was,  however,  to  disappear  temporarily  by  January  1929:  with  the 
exception  of  Ford,  all  automobile  producers  then  decreased  their  pur- 
chases of  steel.2  In  November  total  building  fell  off  more  than  seasonally. 
But  barring  building  and  production  of  building  materials,  which  con- 
tinued to  decline,  most  lines  of  industrial  and  commercial  activity  sur- 
passed 1928  output  figures  during  the  first  six  months  of  1929  at  (from 
September  1928)  falling  prices  but  high  profits.  Also  plant  construction 
and  financing  seemed  to  have  taken  out  a  new  lease  on  life.  Quite  a 
list  of  new  things  (at  least  of  the  "induced"  type)  were  being  inaugurated 
in  June,  when  pig-iron  production  reached  a  maximum.  Moreover,  the 

1  This  question  will  be  touched  upon  in  the  next  section.     But  it  is  clear  that  whatever 
it  was  that  may  have  to  be  explained  by  either  the  mania  or  monetary  factors  or  both 
together  superimposed  itself  on,  and  made  "excesses"  of,  "normal"  phenomena  of  the 
same  nature. 

2  The  effect  on  the  steel  industry  was,  however,  mitigated  by  a  simultaneous  increase 
of  demand  from  the  oil  industry. 


1919-1929  793 

aircraft,  radio,  refrigerator  industries  prospered.  So  did  automobiles, 
tires,  machine  tools  and  other  implements,  hardware,  cotton,  silk,  rayon, 
and  cigarette  production.1  The  Kettleman  Hills  oil  field  was  discovered. 
Mail-order  sales  were  running  far  above  the  1928  level.  Department- 
store  sales  reached  a  peak  in  September.  Extra  dividends  were  paid 
in  the  last  week  of  August  by  oil,  chain-store,  mail-order,  steel,  and  flour 
concerns.  The  agricultural  situation  became  a  matter  of  serious  con- 
cern. The  United  States  Treasury  was  paying  5^g  per  cent  in  June; 
federal  reserve  rediscount  rates  rose  to  6  per  cent  by  August. 

Although  there  is  a  valid  objection  to  any  such  statement,  we  may 
take  April  to  mark  the  peak  of  that  (Kitchin)  prosperity.  But  even  the 
inadequate  description  presented  is  sufficient  to  show  that,  whatever 
may  have  been  wrong  in  the  financial  sector,  the  great  divisions  of  indus- 
try and  commerce  either  expanded2  or  contracted — steel,  motorcars, 
building — in  a  perfectly  orderly  way  during  the  subsequent  months 
through  September.  It  is,  therefore,  understandable  that  when  the 
stock  market — not  altogether  unexpectedly — collapsed,  this  did  not 
cause  paralysis  or  even  particularly  strong  pessimism  in  the  business 
world.  What  immediately  happened  was  in  fact  not  much  more  than 
was  foreseen,  viz.,  a  drastic  reduction  in  the  demand  for  "luxuries," 
of  which  speculative  gains  in  stocks  had  been  a  most  important  feeder. 
The  repercussions  of  this  were  expected  to  induce  and  did  induce  con- 
traction all  round,  but  with  money  rates  failing  to  rise  to  panic  figures — 
as  compared  with  their  reaction  in  prewar  crises — improvement  was 
confidently  predicted  for  the  first  half  of  1930.  Among  characteristic 
reactions  of  "big"  business  we  may  note  that  Ford  announced  a  sub- 
stantial reduction  in  prices,  that  United  States  Steel  and  American  Can 
declared  extra  dividends,  and  that  prevailing  opinion  was  strongly 
against  a  decrease  in  wages.  The  withdrawal  of  foreign  funds,  the 
agrarian  situation,  and  such  liquidations  of  concerns  as  occurred3  were — 
quite  correctly — not  considered  decisive. 

It  is  of  the  utmost  importance  to  realize  this:  given  the  actual  facts 
which  it  was  then  possible  for  either  businessmen  or  economists  to  observe, 
those  diagnoses — or  even  the  prognosis  that,  with  the  existing  structure 
of  debt,  those  facts  plus  a  drastic  fall  in  price  level  would  cause  major 

1  The  following  industries  registered  gains  in  employment  in  July :  slaughtering,  ice 
cream,  flour,  pipes,  structural  ironwork,  machine  tools,  furniture,  shipbuilding,  oil,  fer- 
tilizers, boxes,  cement,  electrical  machinery,  rayon,  radio.     Machine-tool  orders  reached 
their  peak  in  May.     Value  of  exports  rose  to  March. 

2  Some  of  them — tin,  rayon,  and  paper,  for  instance — did  so  vigorously  in  the  third 
quarter,  a  few  even  in  the  last  week  of  the  year. 

3  They  were — in  industry  and  commerce — anything  but  sensational.     In  the  last  week 
of  the  year,  for  instance,  the  more  important  ones  were  those  of  a  South  Carolina  cotton 
mill  and  of  the  Laconia  Car  Company.     A  securities  company  also  failed  during  that  week. 


794  BUSINESS  CYCLES 

trouble  but  that  nothing  else  would — were  not  simply  wrong.  What 
nobody  saw,  though  some  people  may  have  felt  it,  was  that  those  funda- 
mental data  from  which  diagnoses  and  prognoses  were  made,  were 
themselves  in  a  state  of  flux  and  that  they  would  be  swamped  by  the 
torrents  of  a  process  of  readjustment  corresponding  in  magnitude  to 
the  extent  of  the  industrial  revolution  of  the  preceding  30  years.  People, 
for  the  most  part,  stood  their  ground  firmly.  But  that  ground  itself 
was  about  to  give  way. 

F.  The  Behavior  of  Systematic  Series  from  1919  to  1929. — Discussion 
will  conveniently  begin  with  three  of  the  four  series  which  enter  into 
our  Pulse  Charts  (I).  Then  we  shall  discuss  the  behavior  of  (outside) 
debits,  incomes,  and  so  on  (II),  and  finally  some  of  the  series  which 
reflect  the  processes  of  the  spheres  of  banking  (III),  speculation  (IV),  and 
monetary  management  (V). 

I.  It  should  once  more  be  emphasized  that  speaking  of  phases  of  an 
incomplete  Kondratieff  involves  a  hypothesis  which  the  future  course  of 
things  might  easily  fail  to  bear  out,  even  if  this  future  were  less  likely 
than  it  is  to  be  dominated  by  external  factors  or,  if  some  of  us  resent 
application  of  this  term  to  government  action,  by  public  expenditure, 
control,  and  planning.  However,  for  the  period  under  discussion  that 
hypothesis  may  be  checked  statistically,  as  in  the  preceding  section  it 
has  been  checked  historically,  by  confronting  with  the  actual  behavior  of 
time  series  the  expectations  which  follow  from  that  hypothesis.  During 
a  Kondratieff  recession  of  which  the  years  from  1919  to  (the  fall  of)  1925 
form,  according  to  the  schema,  a  part,  and  during  a  Kondratieff  depres- 
sion to  which  the  remaining  years  belong,  we  on  the  whole  expect  output 
to  increase  strongly,  more  strongly  than  in  the  preceding  Kondratieff 
prosperity;  price  level  and  interest  rates  to  fall;  and  balances  to  increase 
but  (barring  government  or  central  bank  action  and  autonomous,  or  at 
any  rate  external,  changes  in  the  quantity  of  monetary  gold)  less  strongly 
than  output  and  also  at  a  rate  smaller  than  that  at  which  they  increased 
in  the  previous  upswing.  Some  explanations  and  qualifications  will  be 
added  in  each  case.  But  a  difficulty  which  has  been  encountered  already 
must  be  mentioned  again.  Since  there  is  no  way  of  isolating  the  effects 
of  our  process  and  since  we  can  do  but  little  more — a  little  more  we  can 
do1 — than  indicate  the  direction  of  the  movements  it  tends  to  bring 
about,  we  cannot  numerically  determine  the  extent  to  which  our  process 
has  been  deflected  by  influences  external  to  it,  especially  by  influences 

1  Comparison  with  the  rates  of  change  that  obtained  at  comparable  junctures  in  the 
past  is  not,  however,  a  method  that  can  be  trusted.  We  have  only  two  such  analogous 
cases.  And  each  of  them  displays  peculiarities  of  its  own.  Even  if  that  were  not  so, 
there  is  no  warrant  for  expecting  that,  for  example,  the  rates  of  increase  in  output  and  of 
decrease  in  price  level  should  be  the  same  in  each  instance. 


1919-1929 


795 


which  acted  on  some  element  of  the  system  in  the  same  direction  as  that 
process  itself. 

a.  1.  Our  expectations  as  to  the  behavior  of  Total  Physical  Output 
or,  if  we  wish  to  exclude  the  influence  of  the  short  fluctuations  in  crops 
that  are  due  to  influence  of  weather,  of  Total  Output  of  Manufactures 
(and  Mining)  call  for  the  following  qualifications.  First,  we  know  that 
they  do  not  apply  to  those  subphases  which  we  call  deep  depressions. 
We  know,  in  the  second  place,  that  on  a  Kondratieff  downgrade  prosperi- 
ties of  the  shorter  cycles  should  display  greater  and  not  smaller  rates  of 


1897      1900  1905  1910  1915  1920  1925  1930 

CHART  XLII. — Industrial  production  (see  Appendix,  p.  1068). 


1935 


increase  because,  the  general  conditions  of  a  Kondratieff  downgrade  set- 
ting the  stage  for  strong  increase  of  output,  this  increase  is  likely  to  show 
with  particular  emphasis  under  the  influence  of  additional  producers' 
expenditure.  Third,  it  must  be  remembered  that  for  Kondratieff  down- 
grades in  general,  but  especially  for  the  one  under  discussion,  our  indices 
are  still  more  than  usually  likely  to  understate  the  rate  of  increase  because 
— not  to  insist  again  on  the  unduly  great  role  which  most  indices  attribute 
to  basic  producers'  goods — increased  supply  of  intangibles  (services  of 
all  sorts),  increased  voluntary  leisure,  economies  in  the  use  of  raw  mate- 
rials and  of  intermediate  products,  and  improvements  in  quality  con- 
stitute, as  we  have  seen,  outstanding  features  of  the  period.  With  these 
provisos  in  mind  we  will  now  inspect  Chart  XLII,  which  presents  the 


796  BUSINESS  CYCLES 

curves  of  industrial  output  for  our  three  countries  from  the  beginning  of 
the  Kondratieff  at  which  they  have  been  made  to  coincide.1 

The  first  thing  that  strikes  us  is  the  extent  of  the  differences  in  both 
the  long-run  and  the  short-run  behavior  of  the  three  curves.  Obviously 
the  war  is  only  in  part  responsible  for  it.  The  United  States  curve, 
keeping  its  more  lively  temperament  throughout — the  reader  will  remem- 
ber that  we  account  for  this  mainly  by  the  pace  of  American  develop- 
ment— shows  the  Kitchins  with  a  clearness  that  leaves  nothing  to  desire. 
War  expenditure  merely  accentuated  their  positive  phases  at  the  expense 
of  the  negative  ones.  The  Juglars  could  not  be  read  off  correctly, 
although  we  can  trace  them  if  we  interpret  the  curve  in  the  light  of  busi- 
ness history.  War  effects  make  it  difficult  to  test  our  Kondratieff 
expectation.  A  linear  trend  drawn  through  the  items  from  1898  to 
1912  would,  however,  display  a  smaller  gradient  than  would  a  trend  drawn 
through  the  items  from  1922  to  1929  which,  considering  that  we  are 
using  a  log  scale,  may  perhaps  be  held  to  spell  acceptable  verification. 
But  we  do  not  insist  on  this.  It  is  more  significant  to  note  that  from 
1899  to  1912  output  of  manufactures  increased  at  most  by  between  70 
and  80  per  cent  and  from  1912  to  1929  by  at  least  between  135  and  145 
per  cent  of  the  1899  figure.  Figures  per  head  of  population  suggest  the 
same  conclusion.  According  to  Professor  Mills,  average  annual  per 
capita  increase,  excluding  construction  and  personal  services  directly 
consumed,  was  roughly  1.1  per  cent  for  1901  to  1913  and  2.4  for  1922  to 
1929.2  Inclusion  of  building  would,  of  course,  strengthen  these  indica- 
tions. Census  figures  of  value  added,  corrected  for  price  level,  also 
behave  according  to  expectation. 

It  should  be  observed  that  there  is  nothing  in  this  record  of  postwar 
developments  to  support  a  belief  in  retardation  and,  in  particular,  a 
belief  in  the  restrictive  tendencies  that  are  held  to  be  inherent  in  monopo- 
loid  large-scale  business.  Neither  is  there  in  it  anything  that  could, 
from  experience  with  cyclically  similar  periods  in  the  past,  be  called 

1  Neither  the  curves  on  this  chart  nor  the  output  curves  in  the  pulse  charts  are  strictly 
comparable  (see  Appendix).  In  method  as  well  as  in  quality  of  material  and  in  coverage 
the  indices  differ  to  a  dangerous  extent.  Nor  have  the  League  of  Nations'  efforts  availed 
to  supply  us  with  really  comparable  indices.  See  also  N.  J.  Wall,  Monthly  Index  of  World 
Industrial  Production,  Preliminary  Report,  United  States  Department  of  Agriculture, 
June  1986,  and  G.  P.  Warren  and  P.  A.  Pearson,  The  Physical  Volume  of  Production  in 
the  United  States,  Cornell  University  Agricultural  Experiment  Station,  Memoir  144, 
November  1932. 

2C/.  Economic  Tendencies  in  the  United  States,  Chap.  I,  especially  p.  244.  The 
figures  of  average  annual  increase  for  the  two  periods  are  3.1  and  3.8  per  cent  respectively, 
hence  also  according  to  expectations.  Mr.  Leong  (Journal  of  the  American  Statistical 
Association,  March  1932)  gives  3.5  instead  of  3.8.  It  is  interesting  to  note  that  from  1899 
to  1929  physical  output  increased  by  about  200  per  cent.  Horsepower  installed  in  manu- 
facturing plants  increased  by  about  330  per  cent. 


1919-1929  797 

astounding,  unheard-of,  or  abnormal.1  In  that  respect  the  capitalist 
machine  seems  simply  to  have  been  working  as  it  previously  did  in 
comparable  epochs.  Of  course,  figures  of  total,  or  total  industrial,  out- 
put can  never  prove  anything — either  in  themselves  or  by  comparison 
with  money  incomes — about  the  presence  or  absence  of  "overproduction." 
Aggregates  may,  as  we  know,  hide  any  amount  of  maladjustment.  But 
the  impressive  steadiness  of  the  process2  and  its  perfect  accordance  with 
the  past  is  still  important  to  note. 

The  German  case  hardly  calls  for  comment.  There  is  no  need  to 
elaborate  the  reasons  why  war  demand,  impinging  as  it  did  in  both 
countries  on  the  productive  possibilities  of  a  Kondratieff  recession,  pro- 
duced the  spurt  in  output  which  we  should  expect  from  this  in  the  United 
States  only,  whereas  it  caused  a  slump  in  German  output.  Rebound 
from  the  trough  in  1919  was  first  prolonged  and  then  drastically  reversed 

1  The  rate  of  increase  could  be  made  spectacular  only  by  counting  from  the  trough  of 
1921  or  the  beginning  of  1922  to  the  peak  of  1929,  which  would  be  devoid  of  sense  and,  in 
particular,  distort  the  comparison  with  England.     We  will  notice  at  once  that  this  also 
applies  to  output  per  man-hour.     The  spectacular  rate  of  increase  in  1922  only  made  up 
for  the  low  ones — or  even  backslidings — which  preceded.     It  has  been  rightly  said  that 
industrial  efficiency  sank  to  a  low  level  during  the  war  and  that  American  industry,  too, 
emerged  from  it  with  an  antiquated  and  overstaffed  apparatus,  in  fact  in  a  drowsy  state 
which  illustrates  both  an  important  aspect  of  "inflation"  and  the  function  of  cold  douches 
such  as  the  one  applied  in  1921.     Measuring  "efficiency"  by  deflated  national  income 
divided  by  hours  of  work  done,  we  get  a  minimum  for  1917.     For  1923  to  1929  Professor 
Mills  (op.  cit.,  p.  297)  gives  an  average  annual  increase  in  output  per  wage  earner  in  62 
manufacturing  industries  of  2.7  per  cent.     For  the  precise  meaning  of  this,  see  footnote  on 
that  page.     The  increase  of  physical  output  itself  during  those  6  years  Professor  Mills 
puts  at  19  per  cent,  the  average  annual  rate  being  2.8  per  cent.     Of  course,  "secular" 
shifts  in  the  economic  organism  partly  show  in  the  sharp  decline  that  occurred  from  1920 
to  1930  in  the  percentage  share  in  the  total  of  persons  gainfully  employed  in  agriculture, 
fishing,  and  lumbering;  in  the  smaller  but  still  significant  decline  of  the  share  of  manufacture 
and  mining;  and  in  the  sharp  increase  of  the  share  of  trade,  transportation,  clerical  work, 
and  services. 

2  Again  the  writer  largely  agrees  with  Mr.  Carl  Snyder's  well-known  views  on  that 
subject.     But  it  is  perhaps  not  superfluous  to  emphasize  once  more  that  an  impression  of 
retardation  such  as  Mr.  A.  F.  Burns  attempted  to  establish  is  perfectly  understandable  on 
statistical  and  other  grounds,  especially  in  a  period  characterized  by  economies  in  the  use 
of  raw  materials,  and  that  an  extrapolation  may  easily  fail  to  be  verified  by  events  owing 
to  the  influence  of  institutional  factors  or  of  the  effects  on  motives  of  the  falling  birth  rate 
or,  in  the  end,  of  saturation.     We  will  use  the  opportunity  to  quote  Mr.  Burns  on  another 
point.     "We  may  therefore  conclude  from  our  analysis  of  American  experience  since  1870: 
first,  that  periods  of  sharp  advance  in  the  trend  of  general  production,  which  are  charac- 
terized invariably  by  considerable  divergence  in  production  trends  [quite  so.  J.  A.  S.I, 
have  been  followed  invariably  by  severe  business  depressions;  second,  that  most  of  the 
business  depressions  of  marked  severity  have  been  preceded  by  a  sharp  advance  in  the 
trend  of  general  production  and  considerable  divergence  in  the  trends  of  individual  indus- 
tries."    (Production  Trends,  p.  251.)     For  Kondratieff  downgrades  at  least,  we  entirely 
agree. 


798  BUSINESS  CYCLES 

(1923)  by  inflation  and  the  events  that  culminated  in  the  Ruhr  invasion. 
What  may  be  termed  the  Dawes  recovery,  which  was  cut  short  by  a  reac- 
tion that  was  intensified  by  fiscal  policy,  and  the  upswing  from  the  middle 
of  1926  on,  which,  as  we  saw,  looks  so  much  like  a  Juglar  prosperity, 
are  well  marked  and,  very  roughly,  in  step  with  American  cycles.  That 
upswing  was,  as  far  as  output  figures  go,  relatively  stronger  than  in  this 
country:  the  Berlin  Institute's  monthly  index  of  industrial  production, 
for  example  (basis  July  1924  to  June  1926),  reached  a  maximum  of 
129.2  in  February  1928.  The  Wagenfiihr  index  of  output  of  manu- 
factures and  mining,  plotted  on  our  chart,  differs  in  details  but  substan- 
tially conveys  the  same  impression.  But  this  meant  little  more  than 
that  the  losses  of  the  war  and  inflation  periods  were  being  made  up  for 
— unlike  the  United  States,  Germany  had  by  1929  gained  but  modestly 
as  compared  with  1913.  Output  per  employee  increased  in  the  aggre- 
gate roughly  by  13  per  cent  from  1925  to  1929. 

As  to  both  facts  and  interpretation  the  British  case  is  more  doubtful. 
Our  graph  shows,  of  course,  decline  during  the  war — understandably 
milder  than  in  Germany — and  the  trough  of  1921,  which  we  interpret 
in  the  same  way  as  the  more  serious  American  one.  We  then  observe 
what  makes  a  good  last  Kitchin  of  the  third  Juglar  and  the  two  first 
Kitchins  of  the  fourth  Juglar,  the  beginning  of  the  latter  being  blurred 
by  the  events  of  1926.  But  the  graph  also  shows  that  the  long-run 
behavior  of  the  British  curve  differs  characteristically  from  that  of  the 
two  others.  This  difference  does  not  date  from  the  war  but  is,  on  the  con- 
trary, just  as  much  or  more  marked  in  the  first  decade  of  the  Kondraticff, 
for  which  it  has  already  been  discussed  (Chap.  VIII).  So  far  we  may 
accept  the  testimony  of  our  graph.  The  post-war  behavior  of  British 
output  is,  however,  inadequately  rendered  by  it  and  has  been  proved, 
by  Professor  P.  Douglas,  Mr.  Colin  Clark  and  other  students,  to  come 
much  nearer  to  American  achievement  that  the  chart  suggests.1  We  may 
confidently  assume  that  British  industrial  production  increased  by  at 
least  20  per  cent  between  the  census  of  1907  and  1924  and  by  at  least 
10  per  cent  between  the  latter  year  and  1929.2  In  support  and  explana- 

1  Both  the  Hoffmann  index  and  the  quarterly  index  of  the  London  and  Cambridge 
Economic  Service,  which  has  also  been  plotted,  are  unsatisfactory  as  to  coverage  and 
attribute  too  much  weight  to  old  and  relatively  declining  industries.  But  for  the  prewar 
period  their  reliability  has  not  been  called  into  doubt  by  later  work,  although  to  some 
extent  the  same  argument  also  applies  to  that  time.  Mr.  Colin  Clark  (Statistical  Studies 
on  Great  Britain,  Economic  Journal,  September  1931)  accepts  Mr.  Rowe's  index  for  1907 
to  1918  and  calculates  from  this  index  and  Trade  Union  unemployment  percentages  an 
index  of  output  per  head  which  for  1918  only  reaches  the  1907  level  again.  To  the  present 
writer  this  estimate  seems  too  low,  but  there  cannot  be  any  doubt  but  that  increase  was 
much  below  the  American  and  German  rates. 

a  Until  1980  a  somewhat  pessimistic  estimate  conforming  to  the  impression  conveyed 


1919-1929  799 

tion,  the  fact  should  be  mentioned  that  the  ratio  of  imported  products 
to  total  production  of  manufacturing  industry  was,  throughout  the 
period,  smaller  than  it  had  been  in  1913 — industry  shifting,  to  some 
extent  at  least,  from  production  for  export  to  production  for  the  home 
market — a  fact  which  stands  out  particularly  clearly  when  figures  are 
corrected  for  those  raw  materials  that  had  to  be  imported  in  both  cases 
(cf.  Colin  Clark,  op.  cit.,  p.  351).  The  improved  picture  still  leaves  us 
with  an  impression  of  inhibited  performance,  but  otherwise  bears  out 
expectation  as  to  the  behavior  of  industrial  output  in  a  Kondratieff 
downgrade. 

2.  Again,  there  is  nothing  suggestive  of  abnormalities  or  new 
problems  in  the  behavior,  relatively  to  each  other,  of  the  aggregates 
into  which  output  of  manufactures  is  usually  divided.  In  order  to 
exemplify  this,  it  will  suffice  to  observe  the  behavior  in  this  country  of 
producers'  and  consumers*  goods,  on  the  one  hand,  and  of  durable  and 
transient  goods,  on  the  other  (Chart  XLITI). 

by  our  chart  very  generally  prevailed,  in  spite  of  certain  objections  that  were  raised  against 
it  (as,  for  instance,  by  Mr.  G.  L.  Schwartz,  Economic  Journal,  March  1929).  Professor 
Bowley  and  Sir  J.  Stamp  (The  National  Income  1924,  1927,  p.  55)  arrived  at  the  conclusion 
"that  real  home-produced  income  per  head  (of  population)  was  very  nearly  the  same  in 
1911  and  1924;  it  is  improbable  that  it  was  any  greater  in  the  latter  year,  and  it  may  have 
been  4  per  cent  less."  And  the  London  and  Cambridge  Economic  Service  Bulletin  for  April 
1930,  stated  that  from  1924  to  1928  output  per  head  (employed)  was  "  stationary  or 
decreasing."  This  was,  on  the  strength  of  new  evidence,  corrected  in  the  Bulletin  for 
June  1930  so  as  to  read  that  there  had  been  during  those  years  an  increase  of  4  per  cent, 
and  another  of  7  per  cent  from  1928  to  1929.  Mr.  Colin  Clark  (op.  cit.,  p.  857)  comparing 
the  value  figures  of  the  1907  and  1924  census  of  the  Final  Product  of  Industry  minus 
Primary  Materials  Purchased  by  means  of  a  more  appropriate  price  index  arrived  at  a 
Net  Output  of  Industry  of  respectively  675  and  820  million  pounds  of  1907  purchasing 
power,  an  increase  of  21.5  per  cent.  Douglas  and  Tolles  (Journal  of  Political  Economy, 
February  1930)  proceeding  by  another  route  even  arrived  at  23.5  per  cent.  As  pointed 
out  by  Mr.  Clark,  it  would  follow  that  real  output  per  worker  increased  by  about  10  per 
cent,  in  spite  of  the  reduction  of  hours.  These  estimates,  so  it  seems  to  the  present  writer, 
cannot  be  far  from  the  truth  although,  for  the  purposes  of  a  comparison  with  the  United 
States,  it  must  not  be  forgotten  that  downward  bias  is  not  altogether  absent  from  American 
indices  either.  For  such  a  comparison  see  A.  W.  Flux,  Industrial  Productivity  in  Great 
Britain  and  the  United  States,  Quarterly  Journal  of  Economics,  November  1933.  For 
the  behavior  of  industrial  output  from  1924  to  1929  both  the  Board  of  Trade  Index  of 
Production  and  the  Annual  Index  of  the  London  and  Cambridge  Economic  Service — the 
latter  after  elimination  of  agriculture  and  dwelling-house  construction  which  it  contains — 
may  be  alternatively  used.  Using  the  former  and  deducting  from  the  figure  of  all  insured 
persons  in  employment  those  engaged  in  building,  transport,  distribution,  and  services, 
Mr.  Clark  (op.  cit.,  p.  360)  calculated  another  increase  of  output  per  head  in  manufacturing 
and  mining  of  10  per  cent  for  1924  to  1929.  Mr.  E.  Devons  (Economic  Journal,  September 
1935)  gives  15.1  per  cent  after  excluding  from  the  employment  figures  all  items  not  covered 
by  the  Board  of  Trade  index.  On  the  London  and  Cambridge  Economic  Service  indices 
see  also  Rowe,  An  Index  of  the  Physical  Volume  of  Production,  Economic  Journal,  June 
1927. 


800 


BUSINESS  CYCLES 


We  find  the  divergences  which  we  should  expect  and,  in  particular, 
we  see  that  the  output  of  consumers'  goods  as  a  rule  (notably  in  1924  and 
from  the  beginning  of  1930  on)  decreases  less  strongly  in  troughs  and 
increases  less  strongly  on  upgrades  than  does  the  output  of  producers' 
goods.  But  we  also  find,  exactly  as  we  did  in  the  case  of  comparable 


1919  1920  1921  1922  1923  1924  1925  1926  1927  1928  1929  1930 1931  1932  1933  1934 
CHART  XL1IL — United  States  production  series  (see  Appendix,  p.  1068). 

prewar  series,  that  the  difference  is,  excepting  the  years  from  1919  to  1921 
and  those  of  the  world  crisis,  neither  so  strong  nor  so  regular  as  some 
theories  suggest.  On  the  whole,  both  series  move  well  together.1  Again, 

1  This  holds  for  other  countries  also.  See,  for  example,  Statistical  Yearbook  of  the 
League  of  Nations,  1931,  Table  91.  In  Germany  output  of  producers'  goods  (1928  =  100) 
was  97  in  1927,  106  in  1929  and  82  in  1930,  output  of  consumers'  goods  (1928  =  100) 
respectively  99,  94,  and  86.  The  fact  that  the  latter  should  have  fallen  so  considerably 
while  the  former  was  still  rising  is  partly  due  to  the  particularly  strong  downward  bias 
of  that  consumers'  good  index  (textiles,  motorcars,  boots,  china,  pianos,  watches  and 


1919-1929  801 

the  really  significant  difference  is  between  durable  and  nondurable  goods. 
Since  the  relative  importance  in  the  consumer's  budget  of  the  former 
naturally  increases  with  increasing  wealth,  we  have  here  a  factor — though 
not  necessarily  a  decisive  one — which  makes  for  increasing  amplitudes  of 
cyclical  fluctuations:  to  take  an  outstanding  instance,  new  passenger 
car  registrations  dropped  from  about  3.8  millions  in  1929  to  1.1  millions 
in  1932. 1  It  is  interesting  to  note  that  while  production  of  durable 
producers'  goods  and  production  of  durable  consumers'  goods  moved 
substantially  parallel,2  yet  the  latter  increased  more  strongly.  The 
average  annual  rate  of  increase  in  the  production  of  industrial  equipment 
— including  nonresidential  construction — was,  according  to  Professor 
Mills'  estimate,  6.4  per  cent  for  1922  to  1929. 

Of  course,  excepting  special  cases,  we  do  not  know  with  any  exactness 
how  much  of  this  was  mere  replacement,  how  much  was  replacement 
plus  improvement,  how  much  was  net  addition,  and  how  much  of  that 
served  new  purposes.  Even  if  we  did,  comparison  of  this  figure  with 
prewar  rates  might  easily  be  misleading,  because  the  period  under  dis- 
cussion happens  to  end  with  a  Juglar  prosperity  and  a  Juglar  recession 
in  which  real  investment  would  be  higher  than  usual.  But  Professor 
Kuznets'  figures  suggest  that  real  investment  was  much  more  modest 
than  is  very  generally  believed3  and  lend  little  support  to  any  of  the 
current  theories  of  the  overinvestment  type.  This,  however,  does  not 
mean  that  production  of  equipment  goods  had  nothing  to  do  with  the 
crisis,  for  it  undoubtedly  reflected  the  process  of  industrial  reconstruc- 
tion, which  was  going  on  at  an  increasing  rate  that  implied  quick  obso- 

paper),  yet  is  not  wholly  spurious.  Investment  in  equipment  industries  was  an  outstanding 
feature  during  the  years  from  1926  to  1929. 

1  Figures  from  R.  L.  Polk  and  Co.,  as  published  in  Automobile  Facts  and  Figures,  1937 
ed.,  by  the  Automobile  Manufacturers'  Association.     There  was,  moreover,  the  shift 
toward  cheaper  cars. 

2  Together,  durable-goods  industries  of  both  types  employed,  in  the  average  of  1921  to 
1930,  about  52  per  cent  of  all  factory  workers  (estimate  by  Colonel  Ayres). 

3  See  S.  Kuznets,  National  Income  and  Capital  Formation,  1919-1935  (1938),  Table  13, 
p.  48.     The  relevant  item  is  II,  1,  6,  net  capital  formation  by  business  exclusive  of  net 
changes  in  inventories  (which  are  of  a  special  nature  and  should  not  be  lumped  together  with 
plant  and  equipment;  the  reader  will  furthermore  recall  the  reasons  why  we  exclude 
residential  building  and  "public  capital  formation,"  excepting  productive  establishments 
owned  by  public  bodies).     At  1929  prices  the  average  of  1919  to  1929  is  only  2.5  billions. 
But  owing  to  the  course  of  the  cyclical  phases,  the  series  moves  successively  on  three 
different  levels:  around  an  average  of  1  billion,  1919  to  1922;  then  for  2  years  at  2.3  billions 
per  year,  while  for  1925  to  1929  the  average  is  3.7  billions.     This  also  shows  (the  1929 
figure,  4.3  billions,  is  the  highest)  that  there  was  no  slackening  in  real  investment  previous 
to  and  independent  of  the  crisis  itself.     The  employment  index  for  the  machine-tool 
industry  (1926  =  100)  gives  the  same  impression.     It  was  82  for  1924,  85.8  for  1925,  100 
for  1926,  92.8  for  1927,  100,8  for  1928,  and  129.8  for  1929. 


802  BUSINESS  CYCLES 

lescence  and  was  bound  to  enforce  drastic  adjustments  and  to  pass  death 
sentence  on  many  plants  and  concerns.  That  process  is,  as  we  know, 
the  very  soul  of  every  recession  or  depression.  Some  of  these  adjust- 
ments, in  the  motorcar  industry  for  instance,  were  carried  out  currently. 
Others  were  not,  and  from  this  source  flowed  what  may  indeed  be 
called  an  accumulation  of  maladjustments,  though  it  would  be  wrong  to 
attribute  it  to  an  overproduction  of  equipment.  Extensive  liquidation 
became  necessary — not,  however,  in  order  to  reestablish  an  equilibrium 
disturbed  by  overexpansion  in  aggregates,  but  in  order  to  reestablish  an 
equilibrium  disturbed  by  innovation  the  roots  of  which  dated  back  to 
the  nineties.  This,  of  course,  formulates  the  matter  in  terms  of  the  long 
wave  only,  but  the  influence  of  the  Juglars  and  Kitchins  can  easily  be 
inserted. 

Other  evidence  strengthens  the  case  for  interpretation  in  our  sense. 
Among  other  things,  the  attitude  of  businessmen  accords  very  well  with 
it.  As  pointed  out  in  the  preceding  section,  we  find  throughout  the 
twenties  and  in  the  midst  of  prosperities  businessmen  struggling  against 
the  results  of  their  own  actions — those  phenomena  repeatedly  noticed 
in  this  book  (for  example  in  Chap.  Ill)  which,  ranging  from  talks  about 
overproduction  and  lack  of  purchasing  power  to  support  given  to  cer- 
tain types  of  corporative  action  and  planning,  were  eventually  to  find 
expression  in  the  NRA  legislation,  the  idea  of  which  characteristically 
enough  originated  in  business  circles.  This  is  precisely  what  we  should 
expect  to  observe  in  the  course  of  the  competing-down  process.  It  is  a 
question  of  some  interest  whether  these  tendencies  were  any  weaker  in  the 
twenties  and  seventies  of  the  nineteenth  century.  Certainly,  they  met 
with  less  success  than  they  eventually  did  in  our  own  time.  Capitalist 
governments  and  parliaments  were  much  less  anxious  to  guarantee 
profits  and  to  protect  from  losses  than  their  less  capitalist-minded 
successors.  And  there  is  really  no  paradox  in  this. 

Moreover,  throughout  the  twenties  and  even  in  the  heyday  of  pros- 
perity phases  and  high  profits  there  was  considerable  excess  capacity1 

1  It  has  been  pointed  out  before  that  excess  capacity  in  any  relevant  sense  is  extremely 
difficult  to  establish  and  to  measure,  and  why  this  is  so.  The  work  by  Nourse  and  Asso- 
ciates, America's  Capacity  to  Produce,  has  been  severely  criticized.  But  it  seems  per- 
missible to  say  that  the  evidence  there  presented  suffices  to  support  the  statement  in  the 
text.  The  same  may  be  said  of  the  study  of  a  number  of  industries  by  R.  F.  Martin,  Bul- 
letin of  the  Taylor  Society,  June  1932:  nearly  every  one  of  the  industries  covered  displayed 
underutilization  of  capacity  in  every  year,  1921  to  1932.  The  question  was  also  investi- 
gated by  a  subcommittee  of  the  Columbia  Commission  on  Economic  Reconstruction  (see 
Report,  Appendix  I).  First,  that  committee  was  permitted  to  use  the  results  of  a  study 
by  L.  P.  Alford  and  J.  E.  Hannum  on  four  industries,  and  second,  it  issued  a  questionnaire 
of  its  own  to  a  considerable  number  of  engineers  and  executives.  Results  are  more  relevant 
to  efficiency  (as  measured  by  product  per  man-hour)  than  to  degree  of  utilization  and  give 


1919-1929  803 

which  was  not  confined  to  ailing  industries  or  regions  nor  entirely  explic- 
able on  oligopolistic  strategy.  The  reader  knows  how  many  funda- 
mentally different  situations  are  covered  by  that  term  and  it  is  clear 
that  instances  of  all  of  them  could  easily  be  adduced.  The  one  that  is 
relevant  here  consists,  first,  in  the  presence  of  old  plant — it  need  not  be 
old  in  years — simultaneously  with  new  plant  and,  second,  in  the  fre- 
quent necessity  of  erecting  plant  and  equipment  on  a  scale  intended  to 
provide  for  future  developments  of  demand.  As  we  have  seen  before, 
the  more  progressive  an  economic  system  is,  the  more  of  these  types  of 
excess  capacity  it  must  display — and,  also,  would  have  to  display  in  a 
socialist  state.  This  has  palpably  been  the  case  in  this  country  during 
the  twenties.  There  are  few  industries,  such  as  oil  and  shoes,  in  which 
presence  of  plant  not  used  up  to  economic  optimum  is  clearly  of  another 
type,  i.e.,  not  incident  to  some  process  of  change.  The  usual  reasoning 
about  excess  capacity  in  general,  therefore,  entirely  misses  the  salient 
point,  but  the  relevance  of  the  facts  so  designated  to  the  explanation  of 
the  subsequent  crisis  becomes  only  the  more  obvious.  Planned  resistance 
by  producers  is,  of  course,  likely  to  accentuate  consequences.1 

3.  Finally,  like  excess  capacity,  supernormal  unemployment  is — pre- 
cisely in  association  with  strong  increase  in  output — an  essential  element 
of  our  picture  of  Kondraticff  downgrades.  We  have  observed  it — each 
time  in  that  characteristic  association — in  the  two  previous  instances, 
and  we  find  it  in  this.  The  postwar  unemployment  is,  however,  a  com- 
plex phenomenon  and  will  have  to  be  touched  upon  again  below.  Here 
it  is  only  the  technological  or  rationalization  component  which  calls  for 
notice.  Some  facts  and  figures  about  product  per  man-hour  or  wage 
earner  have  been  mentioned  already,  which  also  show  once  more  how 
misleading  must  be  any  theoretical  schema  that  assumes  proportionality 
between  output  and  employment  for  any  period  longer  than  a  Kitchin. 
In  further  illustration  we  will  glance  at  a  few  data  taken  from  the  investi- 
gation of  the  Berlin  Institute  on  unemployment  in  Germany.2  Product 
per  workman  increased  from  1926  to  1929  (1930)  by  25  per  cent  in  the 

rise  to  various  doubts.  It  is,  however,  interesting  to  note  how  (relatively)  often  the 
question  what  factors  had  been  responsible  for  keeping  output  below  the — widely  differing 
—estimates  of  possible  maxima,  has  been  answered  by  "competition  and  substitution." 
As  might  be  expected,  lack  of  effective  demand  was  the  most  frequent  answer  and  the  one 
which  seems  particularly  to  have  impressed  the  committee.  In  view  of  the  inference 
drawn  from  it,  it  is  perhaps  not  superfluous  to  state  that  it  does  not  follow  that  inadequate 
money  incomes  were  the  villains  of  the  piece. 

1  "For  we  think  that  the  violence  of  the  convulsions  such  as  recurred  1907-1908, 
1920-1921,  1929-1933  is  due  largely  to  the  partial  character  of  the  liquidation  effected 
during  mild  contraction."     Professor  W.  C.  Mitchell  in  National  Bureau  of  Economic 
Research,  Bulletin  61,  for  Nov.  9,  1936. 

2  A.  Reithmger,  Stand  und  Ursachen  der  Arbeitslosigkeit  in  Deutschland,  1932. 


804  BUSINESS  CYCLES 

iron,  steel,  metal,  machine,  and  motorcar  industries;  by  18  per  cent  in 
mining;  by  16  per  cent  in  the  woodworking  industries;  by  15  per  cent  in 
building;  by  13  per  cent  in  the  chemical  industry;  by  10  per  cent  in  the 
paper,  printing,  food,  and  textile  industries;  by  5  per  cent  in  the  leather 
industry.  Comparable  figures  of  employment  are  not  available  in  all 
cases.  But  we  know,  for  instance,  that  in  the  production  of  crude  steel 
employment  fell  between  1925  and  1927  by  about  10  per  cent;  in  coal 
mining  between  1925  and  1929  by  over  5  per  cent,  in  the  motorcar  indus- 
try also  by  about  5  per  cent;  while  in  the  machine,  paper,  printing,  elec- 
trical, chemical,  building,  food,  textile  and  clothing,  woodworking,  and 
ceramic  industries  there  was  (mostly  modest)  increase.  Fuller  data  for 
output  per  wage  earner  or  even  per  man-hour  in  individual  industries  are 
available  for  this  country.  Reference  should  again  be  made  to  Professor 
Mills's  Economic  Tendencies.1 

But  while  increase  in  product  per  man-hour  may  be  used  to  illustrate 
one  aspect  of  that  period  of  rationalization,  it  is  not  superfluous  to  repeat 
that — irrespective  even  of  purely  statistical  questions  that  may  make 
all  the  difference  in  results — it  must  for  the  purpose  in  hand  be  used 
with  more  caution  than  is  usually  applied.  It  does  not,  of  course,  either 
isolate  or  measure  the  technological  component,  if  indeed  the  word 
component  be  permissible  at  all,  of  total  unemployment  or  the  effect  of 
innovation  on  total  output.  It  is  a  gross  error  to  look  upon  change  in 
product  per  man-hour  and  change  in  total  production  as  two  independent 
factors  militating  against  each  other  and  producing  actual  employment 
as  their  resultant.  Moreover,  nothing  about  "secular"  trends  or  tend- 
encies inherent  to  the  system  can  be  inferred  from  the  survey  of  the 
facts  of  so  short  a  period,  in  which  ultimate  results  cannot  possibly  have 
had  time  to  show  their  true  faces  and  which  was  one  of  rapid  readjust- 
ments. Nevertheless,  we  may  confront  the  impression  conveyed  by 
figures  about  man-hour  product  with  the  behavior  of  employment  on  the 
understanding  that  we  are  dealing  with  but  two  of  many  interdependent 
variables  and  that  no  conclusions  about  causal  relations  can  be  derived 
in  this  way. 

Total  employment  increased,  of  course,  in  all  three  countries.  Part- 
time  employment — to  which  it  would  strictly  be  necessary  to  add  uneco- 
nomic employment,  especially  in  agriculture — changes  in  school  age  and 
in  enforcement  of  attendance  at  school,  in  the  employment  of  women,  in 
age  distribution,  for  Germany  also  the  abolishment  of  compulsory  mili- 
tary service  and  partial  annihilation  of  the  rentier  class  not  only  make 
comparison  with  changes  in  population  more  difficult,  but  also  deprive 
it  of  much  of  its  meaning.  But  it  may  be  stated  that  available  indica- 
tions for  this  country  and  for  Great  Britain  point  to  the  conclusion  that 

1  See  in  particular  pp.  296  and  297. 


1919-1929  805 

total  employment  increased  from  the  end  of  1922  to  the  end  of  1929  in 
such  a  way  as  to  produce  no — or,  if  the  reader  prefer,  a  substantially  hor- 
izontal— "descriptive  trend"  in  unemployment  percentage.  Although 
that  level  was,  as  we  should  expect,  considerably  higher  than  during  the 
Kondratieff  prosperity,  there  must,  hence,  have  been  "absorption  of  the 
technologically  unemployed"  and  of  the  increase  in  employable  popula- 
tion within  that  period  in  this  sense,  although  in  no  other.1 

Of  course,  this  does  not  apply  to  any  individual  industry  or  group  of 
industries.  Manufacturing,  or  manufacturing  and  mining,  forms,  for  the 
purpose  in  hand,  an  arbitrary  group.  Also,  since  it  includes  the  most 
"progressive"  branches  of  economic  activity,  it  is  not  a  random  sample. 
Hence,  no  conclusions  can  be  drawn  from  its  conditions  alone.  In  all 
three  countries  employment  in  manufacturing  industries  increased  but 
slightly,  if  at  all.  In  Great  Britain  the  number  of  persons  employed  in 
manufacturing  and  mining  increased  from  a  little  below  6  millions  in  1924 
to  a  little  over  6  millions  in  1929,  while  the  corresponding  number  for 
building,  transportation,  distribution,  and  services  increased  by  600,000. 
In  Germany  industrial  employment  according  to  the  Institut  fur  Kon- 
junkturforschung  increased  from  1925  to  the  middle  of  1929  by  about 
^  million,  or  5  per  cent.2  In  this  country  census  figures  reflect  the  great 
increase  that  occurred  during  the  war  period  in  the  number  of  wage 
earners  employed  by  the  manufacturing  industries  covered.  This 
increase  was  practically  wiped  out  between  1919  and  1921,  but  nine- 
tenths  of  the  loss  of  those  years  was  recovered  from  1921  to  1923.  The 
next  two  biennial  periods  showed  a  loss  of  about  470,000,  the  period 
from  1927  to  1929  a  gain  of  about  360,000.  Comparison  via  census 
pay-roll  figures  gives  between  1919  and  1929  a  loss  of  about  160,000,  while 
by  extending  coverage  the  loss  from  1923  to  1929  can  be  converted  into  a 
small  gain.3  Over  the  period  there  was  thus  substantial  constancy 

1  No  such  statement  can  be  made  about  Germany  because  of  the — "stabilization" — 
peak  in  unemployment  which  occurred  in  the  winter,  1923-1924.     If  we  exclude  it,  then 
the  "trend"  will  depend  on  our  choice  of  a  precise  spot  between  that  peak  and  the  minimum 
of  the  summer  1925.     From  the  minimum  in  1927  to  the  end  of  1929  the  gradient  of  the 
"trend"  would  be  strongly  positive,  but  that  does  not  mean  much  especially  as  it  may  be 
accounted  for  by  circumstances  not  directly  relevant  to  our  present  subject. 

2  Reduced  to  comparability,  however,  that  figure  would,  the  writer  believes,  dwindle 
to  still  more  modest  proportions. 

8  C.  Goodrich  and  Associates,  Migrations  and  Economic  Opportunities,  1936,  puts  the 
increase  from  1928  to  1929  at  185,000.  Professor  Mills,  op.  cit.,  p.  480,  stated  that 
employment  in  all  manufacturing  industries  increased  at  the  rate  of  0.6  per  cent  a  year 
between  1922  and  1929.  But  this  is  as  likely  to  be  misunderstood  as  the  statement  that 
employment  "declined"  from  1923  to  1929  (see  Aspects  of  Manufacturing  Operations, 
National  Bureau  Bulletin,  May  1935).  Mr.  D.  Weintraub  (Displacement  of  Workers 
by  Increase  in  Efficiency  and  their  Absorption  by  Industry,  1920-1931,  Journal  of  the 
American  Statistical  Association,  December  1932)  found  that  no  "permanent"  displacement 


806  BUSINESS  CYCLES 

partially  reflected  in  the  drastic  fall  in  labor  cost  per  unit  of  product 
in  the  manufacturing  industries  from  1921  on  (see  Mills,  op.  cit.t 
p.  413.)  This  does  not  prove,  of  course,  that  developments  in  manu- 
facturing did  nothing  toward  providing  employment  for  the  increasing 
population,  because  the  sources  of  employment  which  took  care  of  the 
increment  in  labor  supply  were  directly  and  indirectly  created  by  it. 
Nor  does  it  prove  that  the  unemployment  of  the  period  can  be  exclusively 
described  in  terms  of  the  technological  factor. l  But  it  does  prove  that  the 
latter  actually  did  play  the  role  which  we  attribute  to  it  in  pur  picture 
of  Kondratieff  downgrades. 


can  be  proved  for  the  period  1920-1926,  greater  numbers  having  been  absorbed  "through** 
increase  of  output  than  were  displaced  "by"  increase  in  "efficiency,"  while  1926-1929  only 
about  0.1  per  cent  of  the  workmen  displaced  failed  to  be  absorbed,  but  that,  of  course, 
temporary  displacement  was  considerable — about  %  million  per  year.  See,  however, 
J.  Lubin,  Absorption  of  the  Unemployed  by  American  Industry,  Brookings  Institution, 
1929.  On  the  whole,  those  results  do  not  seem  to  diverge  from  either  our  impression  as 
to  the  facts  or  our  interpretation.  Much  information  that  is,  in  spite  of  doubts  on  statis- 
tical and  theoretical  grounds,  very  valuable  has  been  forthcoming  of  late  as  a  result  of  a 
National  Research  Project  of  the  WPA.  See,  in  particular,  D.  Weintraub  and  H.  L. 
Posner,  Unemployment  and  Increasing  Productivity,  March  1937.  We  will  mention 
only  a  few  figures  about  the  system  as  a  whole,  without  attempting  to  appraise  their  precise 
significance.  Taking  1920  for  base,  the  authors  give  146  as  index  of  total  physical  national 
product  for  1929,  126  as  index  of  product  per  man-year,  and  118  as  index  of  total  available 
(usable)  labor  supply.  It  would  follow  that  during  that  period  the  system  as  a  whole 
"absorbed"  much  more  than  the  technological  unemployment  it  "created,"  in  fact,  almost 
that  unemployment  plus  the  simultaneous  increase  in  available  supply  of  labor — a  result  in 
substantial  accord  with  such  other  evidence  as  we  have.  Such  statements  are,  however, 
open  to  a  well-nigh  innumerable  host  of  objections,  and  certain  obvious  inferences  cannot 
fail  to  rouse  the  ire  of  the  more  ardent  critics  and  defenders  of  capitalism.  The  ones  will 
object  that  for  many  workmen  what  looks  statistically  to  be  temporary  loss  of  employment 
often  is  permanent,  that  reemployment  is  very  often  secured  after  delay  spelling  much 
hardship,  that  even  when  secured  the  new  job  may  be  transitory,  less  skilled,  or  otherwise 
less  desirable,  that  the  "services  •  •  •  are  merely  a  buffer  margin  to  enable  the  present  sys- 
tem to  frustrate  its  own  genius  in  the  interest  of  its  creditors"  (Mr.  A.  McLeish,  Machines 
and  the  Future,  The  Nation,  Feb.  8,  1938),  and  so  on.  The  others  will  point  to  the  steady 
and  considerable  rise  in  total  money  and  real  pay  rolls  which  is  relevant  to  the  question, 
not  only  sub  specie  of  a  "compensation"  to  the  working  class  as  a  whole,  but  also  because 
it  does  not  necessarily  follow  that  displacement  would  have  been  as  widespread  if  wages 
had  increased  less  (see  below);  to  long-run  tendencies  such  as  are  displayed  by  the  net 
increase  in  jobs  since  1900;  to  the  arbitrariness  of  a  standard  which  takes  the  superiority 
of  alternatives  for  granted,  and  so  on.  Any  social  system  stands  a  priori  condemned  by 
some  and  a  priori  justified  by  others.  But  there  is  not  much  room  for  disagreement 
about  the  facts. 

1  No  unemployment  ever  can,  except  in  the  shortest  of  runs.  When  Professor  J.  M. 
Clark  (Economics  of  Planning  Public  Works,  National  Planning  Board,  1935)  testified 
to  his  belief  that  "mere  technological  progress  seems  capable  •  •  •  of  bringing  on  a  state 
of  chronic  inability  to  use  all  our  labor  power,'*  he  wisely  inserted  "lacking  the  necessary 
adjustments."  With  this,  to  be  sure,  everybody  will  agree. 


1919-1929  807 

b.  Next,  turning  to  price  levels  in  order  to  see  how  far  the  facts  lend 
themselves  to  interpretation  in  the  light  of  our  model,1  we  must  first 
qualify  our  expectation  for  Kondratieff  downgrades  by  taking  account  of 
the  influence  of  Juglar  and  Kitchin  phases.  Then  we  must  bear  in  mind 
that  price  indices  will  understate  the  real  decline,  for  the  same  reasons 
why  quantity  indices  understate  the  real  increase  plus  the  additional 
one,  that  in  some  cases  the  prices  actually  paid  are  lower  than  the  quota- 
tions which  enter  into  the  indices.  Finally,  we  have  to  add  the  second 
great  factor  that  made  for  decline,  the  reaction  of  the  system  to  the  war 
disturbance. 

What  we  have  before  us  is  obviously  the  combined  effect  of  both. 
Even  in  theory,  let  alone  numerical  operation,  it  would  be  extremely  dif- 
ficult, if  not  impossible,  to  divide  it  up  between  them.  The  two  following 
sets  of  propositions  are  worth  stating,  however. 

First,  if  the  war  disturbance  had  impinged  on  an  otherwise  stationary 
process  and  if  no  permanent  change  had  occurred  in  the  monetary  sphere 
— the  war  being  financed  merely  by  straining  an  existing  and  unchanged 
monetary  and  credit  apparatus — then  prices  would  eventually,  in  the 
course  of  a  process  of  repayment  of  government  debts  from  taxation,  have 
fallen  to  prewar  levels.  Any  permanent  expansion  of  the  sphere  of 
money  and  credit,  especially  if  carried  out  by  all  countries  in  step,  would 
pro  tanto  have  removed  this  mechanical  "deflationary"  pressure.  But 
this  does  not  mean  that,  in  the  absence  of  it,  no  fall  at  all  would  have 
occurred,  for  the  mere  process  of  normalization  in  the  business  sphere,  the 
cessation  of  war  demand,  the  reopening  of  blocked  channels  of  trade,  the 
resumption  of  normal  production  and  habits  would  have  been  sufficient 
to  cause  both  a  downward  jerk  and  a  permanent  lowering  of  price  levels 
from  war  peaks:  the  self-deflation  of  business  cannot  be  prevented  by 
mere  abstention  from  "deflation,"  but  only  by  continuing  "inflation" 
by  additional  government  expenditure. 

Second,  if  there  had  been  no  war  and  if  no  autonomous  change  had 
occurred  in  the  sphere  of  money  and  credit,  then  it  follows  from  our  model 
that  by  about  1925,  when  a  neighborhood  occurred  for  all  three  cycles, 
the  price  level  should  have  been  somewhat  below  that  of  the  preceding 
three-cycle  neighborhood,  viz.,  of  1897.  From  about  1911  the  steady 

1  In  a  sense  this  may  also  be  called  an  attempt  to  delineate,  and  to  compare  with  the 
actual,  the  "  natural"  course  of  events.  But,  if  we  chose  this  mode  of  expression,  we  should 
have  to  recall  once  more  our  distinction  between  equilibrium  influences  and  equilibrating 
influences.  A  priori  there  need  not  be  more  virtue  in  a  "natural"  economic  process  than 
there  is  in  the  natural  course  of  tuberculosis.  And  all  that  we  actually  claim  for  the  former 
is  that  it  fills  certain  "physiological  functions" — a  turn  of  phrase  which  need  not  again  be 
justified — and  cannot  in  genera)  be  interfered  with  without  producing  as  well  as  removing 
undesired  results.  In  this  respect  there  is,  however,  a  great  difference  between  the  reaction 
of  the  system  to  the  war  disturbance  and  its  reacting  to  our  process. 


808  BUSINESS  CYCLES 

pressure  of  " progress" — no  need  to  explain  this  again — would  have 
increased  real  incomes  by  enforcing  a  steady  fall,  which  as  we  know 
should,  in  spite  of  so  much  permanent  expansion  of  the  monetary  sphere 
as  our  process  would  of  itself  entail  (see  Chaps.  IV,  XI,  and  XII),  have 
landed  the  price  level  at  some  figure  below  that  of  the  neighborhood  from 
which  the  current  Kondratieff  started.  If  it  did  not,  this  must  have  been 
due  to  the  war  disturbance  and  other  factors  of  monetary  expansion,  such 
as  the  increase  in  gold  stock  since  the  nineties  or  the  increase  of  lending 
facilities  through  banking  legislation,  both  taking  effect  in  successive 
prosperities.  But  autonomous  change  in  the  sphere  of  money  and  credit 
does  not  per  se  eliminate  the  systematic  tendency  of  price  level  to  fall  in 
a  Kondratieff  downgrade.  It  only  raises  the  level  on  which  this  tendency 
acts.  Hence,  if,  say,  devaluation  had  been  resorted  to  in  order  to  pre- 
vent the  price  fall  of  1920  and  if  it  had  been  successful  in  preventing  it,  the 
period  would  still  have  displayed  that  tendency  and  the  phenomenon 
which  for  us  is  perfectly  natural  but  for  many  economists  a  paradoxon, 
viz.,  prosperity  with  falling  prices.  This  fall  we  would  expect  to  be 
interrupted,  or  temporarily  turned  into  a  small  increase,  around  the 
beginning  of  the  fourth  Juglar,  and  then  to  resume  at  an  accelerating 
rate  as  depression  approached. 

1.  Inspecting  the  curve  traced  out  by  the  price  level  in  this  country 
as  represented  by  the  B.L.S.  index  of  wholesale  prices  (Chart  XXXIX), 
we  first  of  all  notice  the  drop  from  the  (monthly)  maximum  of  167.2 
(1926  =  100)  which  occurred  in  May  1920,  to  a  minimum  of  91.4  in 
January  1922.  This  drop,  or  rather,  if  we  allow  for  the  fact  that  panic 
declines  always  outrun  the  goal  and  that  hence  some  rebound  would  have 
occurred  even  in  the  absence  of  any  other  influence,  a  drop  to  a  few  points 
above  that  figure1  was  due  not  merely  to  normalizing  reaction  to  war 
disturbance:  we  know  that  this  reaction  coincides  with  a  Juglar  depres- 
sion. But  that  reaction  was,  it  is  safe  to  say,  the  dominant  factor  in  it. 
We  also  know  and  shall  definitely  see  presently  that  no  appreciable 
"deflationary"  pressure  was  exerted  on  the  system.  The  huge  machine 
for  credit  creation  set  up  by  the  Federal  Reserve  Act  as  amended  during 
the  war  period  was  left  intact,  and  war  expansion  of  the  credit  structure 
was  supported  and  to  some  extent  camouflaged  by  the  broadening  of 
its  gold  basis.  As  a  postwar  adjustment  that  drop  was,  hence,  almost 
wholly  due  to  the  self  deflation  of  business. 

Now,  that  minimum  of  91.4  was  still  about  30  per  cent  above  the 
annual  average  of  1913,  and  it  is  quite  possible  that  reaction  to  war 

1  This  allowance  is  a  difficult  matter,  for  what  it  is  reasonable  to  attribute  to  rebound 
from  a  trough  exaggerated  by  the  spiral,  is  mixed  up  with  the  effects  of  Kitchin  movements 
and  other  influences  and  cannot  simply  be  read  off  either  from  the  maximum  attained  in 
March  1923  (104.5)  or  from  the  broader  plateau  of  about  99  to  which  price  level  then 
relapsed. 


1919-1929  809 

inflation  continued  to  play  some  role  throughout  the  twenties.  But 
owing  to  the  circumstances  just  mentioned,  this  effect  cannot  have  been 
significant.  At  least  it  is  as  safe  to  say  that  the  rest  of  the  period  was 
dominated  by  other  factors  as  it  was  to  say  that  1920  to  1922  was  dom- 
inated by  that  reaction,  and  it  is  on  this  rest  of  the  period  that  we  have  to 
try  out  our  schema.  Rebound  from  panic  low  plus  the  advent  of  Juglar 
recovery  plus  Kitchin  effect  may  then  be  invoked  in  explanation  of  the 
rise  in  the  B.L.S.  all-commodity  index  to  March  1923,  which  was  mainly 
due  to  the  class  of  semimanufactured  commodities  (128.3  in  April  19231). 
It  was  suspiciously  strong  and,  in  fact,  gave  way  to  relapse  after  the  first 
quarter  of  1923  (annual  figure  for  the  year — monthly  average — 100.6). 
During  1924  there  was  substantial  stability  at  the  lower  level  reached  in 
the  last  three  quarters  of  1923  (annual  figure,  98.1),  in  1925  a  rise  (annual 
figure,  103.5)  followed  by  a  fall  during  1926  (97.9  for  December),  tapering 
off  in  1927,  interrupted  in  1928,  resuming  in  the  last  quarter  of  that  year, 
and  gaining  momentum  in  the  last  quarter  of  1929  (December:  93.3). 
Finished  commodities  fell  practically  steadily  from  the  maximum  of 
102.1  (November  1925). 2  The  index  of  sensitive  prices  came  down  well 
in  1926,  rose  somewhat  in  1927,  sagged  in  1928,  and  slumped  in  the  last 
quarter  of  1929.  Comparison  of  the  "domestic"  price  level  with  those 
of  exports  and  imports3  would  greatly  add  to  the  details  of  the  picture 
without  altering  its  general  contours.  The  National  Industrial  Con- 
ference Board's  index  of  cost  of  living  falls  from  nearly  200  per  cent  of  its 
1914  base  (the  middle  of  1920)  to  a  little  above  160  (the  middle  of  1922), 
rises  to  the  middle  of  1925,  and  then  falls  slowly,  not,  however,  reaching 
160  before  the  middle  of  1930. 

This  is  what  has  been  so  frequently  referred  to  as  the  stability  of  the 
American  price  level4  during  the  twenties  and  taken  as  proof  of  the 

1  This  fact,  however,  would  accord  well  with  the  "rebound  theory,"  since  the  prices 
in  that  class  display  the  highest  of  the  maxima  in  1920  (258  in  May)  and  the  sharpest  fall 
(90.3  in  December  1921). 

2  If  we  eliminate  farm  products,  the  picture  is  from  our  standpoint  greatly  improved : 
we  find  steady  fall  from  1923  on  at  an  average  rate  of  1.5  per  cent  per  year,  varying,  how- 
ever, nicely  in  the  phases  of  the  two  shorter  cycles  (see  Professor  Mills,  op.  cit.t  Fig.  61  on 
p.  341).     The  same  applies  to  processed  producers'  goods  and  from  1925  to  processed 
consumers*  goods,  but  neither  to  raw  producers'  nor  to  raw  consumers'  goods  (see  ibid., 
Fig.  68  on  p.  359). 

3  See,  for  example,  Professor  Theodore  J.  Kreps,  Export,  Import  and  Domestic  Prices 
in  the  United  States,  1926-1930,  Quarterly  Journal  of  Economics,  February  1932.     One 
point  is,  of  course,  obvious,  viz.,  that  export  prices  of  manufactured  goods  were  in  general 
lower  than  the  domestic  prices  of  the  same  goods  and  behaved,  in  their  short-run  tendencies, 
more  according  to  the  competitive  pattern. 

*  It  must  be  remembered  not  only  that  it  was  the  fashion  to  exaggerate  such  stability 
as  there  was  but  also  that  there  was  a  time,  bygone  forever  let  us  hope,  when  many  econ- 
omists believed  that  keeping  the  price  level  constant  is  all  that  is  necessary  in  order  to 
insure  general  stability  of  the  economic  system,  to  avoid  depressions,  etc. 


810  BUSINESS  CYCLES 

success  of  federal  reserve  policy  and/or  of  the  absence  of  "inflation." 
We  shall  return  to  this  later  on.  Meanwhile,  it  is  submitted  that  the 
price  behavior  described  does  not  fail  to  bear  out  our  expectations. 
The  Kitchins  show  well,  the  Juglar  effect  is  not  absent,  and  the  Kondra- 
tieff  pressure  due  to  the  long-term  effects  of  "progress"  is  much  in 
evidence.  In  the  light  of  later  events  and  considering  the  extent  of  the 
expansion  in  physical  product  and  of  the  fall  in  real  cost,  an  impression 
may  dawn  upon  us  to  the  effect  that  the  slightly  inclined  plateau  we 
behold — from  1922  to  1929  the  rate  was  about  %  per  cent  a  year — may 
perchance  not  have  been  all  that  we  should  have  expected  from  the 
unhampered  working  of  our  model,  especially  when  we  weigh  the  inter- 
ruption and  partial  reversal  of  the  falling  tendency  from  the  summer  of 
1927  to  the  summer  of  1929.  And  from  this  we  might  go  on  to  infer 
that  there  was  some  factor  which  prevented  prices  from  falling  as  much  as 
they  would  have  if  left  to  themselves.  But  so  long  as  the  presence  of 
such  a  factor  is  not  proved,  we  have  not  more  than  an  impression  which 
may  easily  mislead. 

2.  The  English  case,  fundamentally  similar  in  other  respects,  differs 
from  the  American  by  the  element  of  significant  pressure  exerted  by  the 
monetary  factor,  which  was  present  in  the  former  and  not  in  the  latter. 
And  since  under  English  conditions  this  pressure  would  primarily  work 
through  affecting  imports  and  exports,  devaluation  of  the  pound  in  1922, 
when  the  price  level  had  had  the  opportunity  of  showing  its  true  face,  to 
62.5  per  cent  of  the  prewar  par,  the  figure  roughly  corresponding  to  that 
of  the  Board  of  Trade  price  index  for  that  year,  would  in  this  particular 
case,  as  has  been  pointed  out  in  Sec.  C,  have  in  fact  removed  a  weight 
that  not  only  depressed  but  also  distorted.  The  Board  of  Trade  index  of 
wholesale  prices  displayed  the  postwar  drop,  falling  from  325  per  cent  of 
the  1913  annual  figure  (April  and  May  1920)  steadily  through  1922 
(December:  156).  Then  it  rose  no  less  steadily  through  1923  and  1924 
(maximum  of  171  was  reached  in  January  1925).  That  both  drop  and 
ascent  lasted  longer  than  in  this  country  is  not  difficult  to  understand. 
But  "overvaluation"  of  the  pound  must  be  invoked  in  explanation  of 
the  subsequent  fall  to  144  in  April  1926.  The  disturbances  of  that  year 
make  it  unsafe  to  speak  of  an  effect  of  a  Juglar  prosperity  being  reflected 
in  the  modest  rise  to  November  (152),  after  which  the  English  level, 
falling  almost  steadily  to  136  for  September  1929,  conforms  to  expectation 
better  than  the  American  one.1  The  Ministry  of  Labour  cost  of  living 
index  (October  1920,  276  per  cent  of  1913;  December  1929,  166)  inter- 
rupted an  otherwise  steady  downward  course  only  in  1923—1924. 

1  But  the  extreme  stability,  at  about  80  per  cent  of  the  1924  level,  of  the  Board  of 
Trade's  index  exclusive  of  foods  during  1927,  1928,  and  three  quarters  of  1929,  somewhat 
militates  against  that  statement. 


1919-1929  811 

3.  The  German  case,  finally,  is  almost  completely  dominated  by  fac- 
tors extraneous  to  our  process.  The  annual  index  of  the  Institut  fiir 
Konjunkturforschung  (1913  ==  100)  resumes  in  1924,  the  first  year  after 
the  gap  caused  by  the  war  and  by  "wild"  inflation,  at  136.  Prices  of 
agrarian  products  were  then  at  112,  sugar  at  176,  products  from  over- 
seas at  125,  and  industrial  materials  at  146. 1  These  figures  obviously 
reflect  past  disturbance,  both  in  their  absolute  values  and  in  their  rela- 
tions to  each  other — the  new  prices  in  terms  of  gold  marks  being  largely 
experimental — and  the  rise  of  the  index  in  the  next  year  and  its  drop  to 
128  for  1926  was  but  the  result  of  adjustments  which  have  little  to  do 
with  normal  cyclical  movements.  From  the  middle  of  1926  to  the 
middle  of  1928  the  general  movement  is  upward  and  thus  contrary  to 
the  English  and  American  tendencies,  though  shorter  fluctuations  arc 
fairly  well  in  step.  This  may,  in  part,  have  been  due  to  the  influence  of 
the  prosperity  phase  of  the  fourth  Juglar,  but  much  more  important 
was  that  of  the  "consumers'  prosperity."  The  influx  of  the  foreign 
credits  that  helped  to  finance  both  was  not,  of  course,  an  additional  factor 
but  only  implemented  the  two  others.  Since,  however,  the  supply  of 
those  credits  constituted  an  independent  element  of  the  situation,  a 
surface  explanation  can  alternatively  be  presented  in  terms  of  them. 
This  will  account  both  for  the  rise  in  prices  in  1925  and  for  their  trough 
in  1926,  but  it  will  not  account  for  the  tapering  off  in  1928  when  beginning 
weakness  of  prices  preceded  the  withdrawals  of  foreign  funds.  We  may, 
moreover,  note  the  heavy  responsibility  of  building  materials  for  the 
rise  of  the  index  in  1927  and  1928  and  the  insignificance  of  the  contribu- 
tion of  nonferrous  metals  and  of  chemicals — facts  which  tend  to  reduce 
the  extent  of  the  abnormality.  The  cost-of-living  index  rose  practically 
without  interruption  from  the  beginning  of  1924  to  March  1929,  and 
thus  presents  the  problem  implied  in  behavior  contrary  to  expectation 
still  more  clearly  than  the  index  of  wholesale  prices.  But  explanation 
by  the  agricultural  policy  and  by  the  "consumers'  prosperity"  also 
becomes  still  more  convincing  when  applied  to  cost  of  living. 

c.  The  attempt  to  interpret  experimentally,  in  the  same  way  as 
price  levels,  the  behavior  of  interest  rates  in  the  light  of  war  disturbances 
and  of  our  model  is  still  more  hazardous,  because  the  influence  of 
"policies"  and  of  other  external  factors — open-market  operations  in  the 
United  States,  the  dominant  role  of  government  financing  and  of  the 
management  of  the  pound  in  England,  all  sorts  of  irregularities  in  Ger- 
many— may  still  more  obviously  than  in  the  case  of  price  levels  seem  to 
preclude  all  reasonable  hope  of  finding  traces  of  our  process.  This  is  a 
question  of  fact,  however,  and  at  the  outset  it  is  sufficient  to  answer 

1  Since  practically  no  finished  products  enter  into  that  index,  we  cannot  trust  it  to 
render  the  real  behavior  of  the  price  level  faithfully. 


812  BUSINESS  CYCLES 

doubts  on  that  score  by  pointing  out  that  policies  and  other  external 
factors  manage  or  influence  money  or  credit  not  in  abstracto  but  in  given 
business  situations  which,  barring  the  theory  that  they  are  nothing  but 
the  products  of  monetary  policy,  must  assert  themselves  both  in  the 
measures  taken  and  in  their  effects. 

Expectation  is  for  a  fall,  possibly  interrupted  by  Juglar  and  Kitchin 
prosperities.  This  applies  to  both  the  war  and  the  cyclical  component. 
We  shall,  as  in  the  case  of  price  levels,  attribute  partly,  though  not 
wholly,  the  peak  in  1920  as  well  as  the  fall  in  1921 — which  in  this  country 
preceded  open-market  purchases  and  may,  hence,  be  looked  upon  as 
"genuine,"  while  the  additional  fall  in  the  first  half  of  1922  raises  another 
problem — to  the  war  and  its  liquidation,  and  apply  expectations  from 
our  model  primarily  to  the  rest  of  the  period,  although,  especially  for 
England  and  Germany,  conditions  traceable  to  the  war  continued  to  influ- 
ence rates  still  more  obviously  than  price  levels.  It  should  be  observed 
at  once  that  our  theoretical  expectation  being  what  it  is,  a  Hayek  effect1 
of  cheap  money  policies  must  be  extremely  difficult  to  discover  and  pre- 
sumably cannot  be  discovered  at  all  from  the  behavior  of  interest  rates 
alone. 

1.  The  strongest  traces  of  what,  from  the  standpoint  of  our  model, 
would  be  normal  behavior  we  shall,  of  course,  expect  to  find  in  the  United 
States.  Although  the  advent  of  the  Federal  Reserve  System  materially 
changed  the  significance  of  the  commercial  paper  rate,2  we  may  still  use 
it  to  represent  the  course  of  short  rates,  especially  as  acceptance,  call, 
and  time  money  rates  would  not  give  significantly  different  results.  If 
we  take  our  stand  at  the  end  of  1921,  when  it  was  at  about  5  per  cent  and 
the  postwar  hump  had  been  a  little  more  than  eliminated,  we  do  in  fact 
find  the  falling  tendency  we  look  for  until  the  end  of  1927  (see  Chart 
XXXIX).  The  trough  in  the  second  half  of  1924  strikes  us  as  abnormally 
deep,  but  the  increase  in  1925  at  the  beginning  of  the  fourth  Juglar  and 
the  substantially  level  movement  during  the  two  following  years  are 
exactly  what  we  would  expect,  especially  if  taken  together  with  the  fall 
that  occurred  in  1930.  But  between  there  is  another  hump:  the  rise  in 
1928  and  to  nearly  the  middle  of  1929  is  distinctly  contrary  to  expecta- 
tion and  calls  for  an  explaining  factor,  which  we  shall  have  no  difficulty  in 
discovering.  An  obvious  clue  is,  however,  afforded  by  the  leadership  in 
that  movement  of  call  rate.  The  Kitchins  show  well. 

1  By  a  Hayek  effect  we  simply  mean  the  effect  on  investment  of  a  rate  of  interest  lower 
than  would  have  obtained  had  the  process  been  left  to  itself.     We  are  not  discussing  the 
applicability  of  Professor  Von  Hayek's  theory  in  its  strict  acceptance,  the  assumptions  of 
which  do  not  seem  to  be  fulfilled  by  postwar  conditions. 

2  The  behavior  of  commercial  paper  rate  in  fact  differs  from  the  prewar  pattern.     The 
seasonal  movement,  for  instance,  has  all  but  disappeared. 


1919-1929  813 

Bond  yields1  illustrate  our  point  still  better  (see  Chart  LIT).  From 
the  end  of  1921  to  the  first  quarter  of  1928  they  (Moody's  index  of  AAA 
bonds)  declined  by  a  little  less  than  1  per  cent  and  this  fall  was,  barring 
the  recovery  from  the  trough  of  1922,  almost  without  interruption. 
Their  increase  from  the  second  quarter  of  1928  to  the  third  quarter  of 
1929  was  but  small  and  is  easily  accounted  for  by  the  abnormal  behavior 
of  short  rates  and  by  the  speculative  mania  in  general.  Though  the 
relation  over  the  whole  period  between  short  rates  and  bond  yields 
differs  from  their  prewar  relation,  the  latter  thus  still  give  the  "trend"  or 
general  drift  of  the  former.  An  impression  to  the  contrary  is  created 
merely  by  the  1924  trough  in  short  rates  not  reproduced  in  long  rates 
and  yields.  Since  the  rise  in  1928  and  half  of  1929  was  also  but  very 
weakly  reproduced  in  yields,  formal  trends  drawn  through  money  rates 
and  yields  from  the  middle  of  1924  to  the  middle  of  1929  would,  of  course, 
show  divergent  movements.  But  there  is  no  justification  for  such  a 
procedure. 

It  has  been  pointed  out  in  Chap.  XII  that  the  rate  on  customers' 
loans  is  not  in  practice  what  it  is  in  the  theoretical  blueprint,  viz.,  the 
core  of  the  interest  structure.  In  Germany,  in  particular,  it  followed 
mechanically  the  Reichsbank  rate,  and  fixing  it  seems  to  have  been 
almost  a  clerical  job.  But  in  this  country  it  still  retained  something  of 
that  pivotal  significance  which  our  schema  assigns  to  it.  Therefore,  it 
seems  worth  our  while  to  glance  at  its  variations  (see  Chart  XLIV). 

1  Reference  is  again  due  to  Mr.  C.  C.  Abbott's  important  studies,  Review  of  Economic 
Statistics,  January  and  May  1935.  It  is  a  matter  of  regret  to  the  present  writer  that  this 
book  had  to  go  to  press  without  the  benefit  it  would  have  derived  from  the  researches  of 
Mr.  F.  R.  Macaulay,  Movements  of  Interest  Rates,  Bond  Yields  and  Stock  Prices  in  the 
United  States  since  1856,  1988.  By  familiar  operations,  which  are  very  simple  in  the  case 
of  the  United  States  but  require  "corrections"  involving  the  exercise  of  much  arbitrary 
judgment  in  the  other  two  cases,  a  rough  covariation  between  the  short-run  variations—- 
mainly the  Kitchins — in  interest  rate  and  in  price  level  could  be  brought  out  graphically. 
We  do  not  do  this  because  all  that  matters  for  the  purposes  of  this  book  is  evident  from 
inspection  of  the  pulse  charts.  It  will  also  be  evident  that  even  in  the  American  case  the 
relation  is  clearly  a  disturbed  one — which  must  be  so  quite  apart  from  external  disturbances, 
since  rates  and  levels  are  not  the  only  elements  of  even  the  aggregative  system — but  it  is 
fairly  close  from  1923  to  1927.  If  the  opposite  "underlying  tendencies"  are  eliminated, 
rough  covariation  is  still  more  obvious  between  interest  and  industrial  output.  In  a  paper 
previously  quoted  (Journal  of  the  American  Statistical  Association,  June  1931,  p.  5)  Mr. 
B.  B.  Smith  compared  the  (smoothed)  business  index  of  the  Cleveland  Trust  Company 
with  the  (smoothed)  deviations  of  a  short  money  index  from  bond  yields  (the  latter  taken 
as  "normal,"  an  idea  which  corresponds  well  with  our  view  of  the  matter)  lagged  by  one 
year,  and  found  an  acceptable  inverse  covariation  between  the  two  through  1928.  A  little 
reflection  should  convince  the  reader  that,  although  we  could  not  stand  for  a  lag  of  exactly 
one  year — but  only  for  a  shorter  one — this  result  lends  itself  to  interpretation  by  the 
Kitchin  phases.  It  should  be  observed  that  it  does  not  contradict  the  "Harvard  lag"  of 
short  rates. 


814 


BUSINESS  CYCLES 


An  average  for  26  to  35  cities  computed  by  the  National  Bureau  of 
Economic  Research  (Bulletin  for  July  18,  1927)  moved  much  like  bond 
yield,  though  more  sluggishly  and  at  a  level  that  was  considerably — by 
about  from  %  per  cent  to  1  per  cent — higher.  Only  in  New  York  City 
rates  on  customers'  loans  moved  until  the  summer  of  1927  at  about  the 
same  level  as  bond  yields,  reflecting  as  they  should  the  rise  of  the  fourth 
Juglar  better  than  these,  but  responding  more  strongly  to  the  abnormal 
conditions  that  obtained  from  the  middle  of  1927  to  the  middle  of  1929. 


1920   I92f     1922    1923    1924    1925    1926     1927   1928     1929    1930    1931     1932     1933    1934 
CHART  XLIV. — United  States;  rates  on  customers'  loans  (see  Appendix,  p.  1069). 

In  the  South  and  West  and  even  in  the  North  and  East  those  rates 
throughout  the  period  stayed  at  a  figure  much  above  Kondratieff  expecta- 
tion, and  if  it  were  not  for  the  fact  that  other  and  much  cheaper  sources 
of  money  were  readily  available  to  all  that  was  most  "progressive'* 
in  the  business  organism  of  those  years,  we  should  be  faced  with  the 
question,  not  of  the  presence  of  a  Hayek  effect,  but,  on  the  contrary,  of 
the  effects  of  dear  money.1  As  it  was,  the  high  level  of  rates  on  customers' 
loans  proves  little  more  than  the  extreme  imperfection  of  the  money 
market  in  our  sense.2 

This  imperfection  and  the  abundance  of  other  sources  of  funds  also 
disrupts  the  close  relation  which  according  to  our  model  should  subsist 
between  customers'  rates  and  profit  ratios  in  Professor  Crum's  sense. 

1  That  question  will  be  taken  up  in  Sec.  F,  III. 

2  The  National  Bureau  average  of  customers'  rates  fell  at  a  rate  of  about  0.8  per  cent 
per  year  from  1922  to  1929,  while  bond  yields  fell  at  a  rate  of  1.4  per  cent.     See  F.  C.  Mills, 
op.  cit.t  p.  455, 


1919-1929 


815 


Inspecting  the  curves  traced  out  by  the  latter  (Chart  XLV),  we  find  but 
unsatisfactory  covariation  with  those  rates.  We  shall  understand  that 
the  trough  in  profit  ratios  that  occurred  in  1921  was  accompanied  by 
high  loan  rates;  we  shall  not  be  surprised  to  observe  the  disruptive  influ- 
ence of  the  speculative  mania  at  the  end  of  the  period;  and  we  may  also 


ALL  SIX  DIVISIONS 


1919  I9ZO  1921   1922  1923  1924  1925  1926  1927  1928  1929  1930  1931  1932  1933 
CHART  XLV. — Profit  ratios  of  United  States  corporations  (see  Appendix,  p.  1069). 

allow  for  lags.     This,   no  doubt,  improves  matters,  but  the  relation 
remains  disappointingly  weak.1 

2.  From  the,  comparatively  speaking,  normal  course  of  interest  rates 
in  the  United  States  the  German  case  indeed  differs,  but  not  so  much  as 
one  might  think.  Descent  during  1924  from  the  impressive  heights  at 
the  beginning  of  that  year — day-to-day  money,  for  example,  was  at 
87.64  per  cent  for  January — has,  of  course,  nothing  to  do  with  our  process 
and  merely  reflects  reaction  to  the  preceding  death  struggle  of  the  specu- 
lation that  was  an  element  of  the  process  of  wild  inflation.  Similarly, 
the  rise  in  1929  only  reflects  the  slipping  away  of  foreign  balances  owing 
to  the  New  York  stock  exchange  boom  and  to  a  wave  of  foreign  distrust. 
But  if  we  connect  (see  Chart  XL)  the  figures  for  the  first  quarter  of 

1  The  covariation  of  customers'  rates  with  the  ratio  of  net  income  to  capital  is  more 
satisfactory,  however. 


816  BUSINESS  CYCLES 

1925  with  the  figures  of  the  first  quarter  of  1929  or,  still  better,  those  for 
the  last  quarter  of  1925  with  those  of  the  last  quarter  of  1930,  we  get  a 
fall  which  may  reasonably  be  attributed  to  our  tendency.  The  fluctua- 
tions within  it,  i.e.,  mainly  the  rise  during  1927  and  the  reaction  to  it, 
also  bear  interpretation  by  the  cyclical  process.  While  it  cannot  be 
urged  against  this  interpretation  that  the  fall  in  1926  was  simply  a 
lagged  effect  of  previous  inflows  of  foreign  funds  and  the  rise  in  19£7 
a  lagged  effect  of  the  temporary  cessation  of  those  inflows  in  1926 — 
because  those  inflows  were  not  independent  of  the  German  business 
situations — it  seems  permissible  to  point  to  the  importance  of  this 
ipethod  of  financing  domestic  business  in  explanation  of  certain  obvious 
irregularities  in  timing.  Of  course,  the  general  level  around  which  rates 
fluctuated,  though  displaying  a  downward  inclination,  was  abnormally 
high  throughout:  6  per  cent  gold  mortgage  bonds  (roughly  though  not 
perfectly  comparable  to  American  triple  A  bonds),  which  were  around 
50  per  cent  of  par  in  1924,  were  still  but  98.95  in  February  1927,  when 
they  began  to  fall  again,  under  the  pressure  of  foreign  sales  and  of  the 
strain  of  the  consumers'  prosperity.  But  this  is  easily  accounted  for  by 
the  wastages  and  dissavings  of  the  war  and  the  period  of  inflation  and  by 
the  fiscal  policy  pursued. 

3.  In  the  English  curve  (Chart  XLI;  for  call  rate  and  the  rate  on 
loans  at  short  notice  to  the  stock  exchange  see  Chart  LIV)  Kitchins  show 
well,  and  there  is  no  reason  for  not  attributing  the  rise  in  1925  partly  to 
the  Juglar  phase  then  prevailing.  But  a  trace  of  the  Kondratieff  tendency 
is,  at  best,  only  visible  from  the  end  of  1925  to  the  end  of  1928.  The 
peak  of  1920  and  the  trough  of  1922  are  again  easily  understandable  on 
short-run  considerations — cyclical  ones  among  them — but  neither  of 
them  can  be  taken  as  a  starting  point.  On  the  other  end,  the  sharp 
rise  in  1929  is  also  easy  to  understand  considering  the  sensitiveness  of  a 
fundamentally  untenable  currency  situation.  We  should  add,  however, 
that  as  shown  by  the  behavior  of  the  rate  on  stock  exchange  loans,  domes- 
tic speculation  was  also  an  important  factor.  But  even  if  we  cut  off 
the  postwar  hump  and  thereupon  connect  the  figure  for  the  first  quarter 
of  1922  with  the  figure  for  the  first  quarter  of  1928,  we  get,  in  contrast 
to  the  American  case,  a  rising  instead  of  a  falling  "trend,"  and  even  the 
trough  of  1930  is  still  above  that  of  1922.  This  is  of  course  due  to  the 
policy  adopted  by  the  Bank  in  protecting  the  pound  after  it  had  climbed 
up  to  par,  and  pro  tanto  reveals  the  presence  of  a  disturbance  of  the 
normal  course  of  events  as  important  as,  though,  of  course,  not  more  so 
than,  anything  that  affects  money  rates  can  be.  Not  until  depreciation 
removed  the  shackles  was  interest  free  to  move  according  to  its  own  law. 

The  high  general  level  of  interest  rates  is,  however,  not  wholly  due 
to  that  but  also  to  causes  similar  to  those  which  were  operative  in  the 


1919-1929  817 

German  case.  The  index  of  yields  on  four  fixed  interest  securities1  fell 
from  its  peak  (November  1920:  163;  1913  =  100)  to  a  trough  in  June 
1923.  Even  at  this  trough  (117)  it  was  perhaps  something  like  30  points 
above  what  it  could  possibly  have  been  without  the  war  and  postwar 
policies.  Variations  as  distinguished  from  general  level  are  nevertheless 
not  so  far  from  expectation  also  in  this  case.  We  should  have  expected 
a  rise  after  that  trough,  owing  to  the  influence  of  Juglar  phases.  We 
find  it,  and  although  it  set  in  earlier  and  went  farther  than  we  might  have 
expected — the  peak  which  occurred  in  September  1926  was  139 — and 
although  other  factors  were  probably  more  important  in  bringing  it 
about,  there  is  again  no  reason  for  not  attributing  it  in  part  to  the  Juglar 
recovery  and  especially  to  the  Juglar  prosperity.  The  slow  fall  to 
January  1929  (122)  also  is  what  we  should  expect  in  a  Juglar  recession. 

The  differences  in  the  behavior  of  interest  rates  in  our  three  countries 
are  not  more  obvious  than  the  parallelisms  between  them.  They  were 
enforced  through  the  mechanism  of  short  balances,  which  for  a  time 
functioned  not  less  but  more  promptly  than  before  the  war.  This  can 
be  seen  from  the  fact  that,  respectively,  from  April  1925  and  August 
1926  until  1930,  the  Berlin  prices  of  sterling  and  dollars  display  very 
close  inverse  associations  with  the  differences  between  the  Berlin  and 
the  London  and  New  York  market  rates.2  These  facts  acquire  for  us 
additional  significance  in  the  light  of  the  other  fact  that  English,  German, 
and  American  cyclical  phases  were  substantially,  though  not  perfectly, 
in  step.3 

II.  a.  For  an  idea  about  the  behavior  of  American  system  expendi- 
ture during  the  postwar  period,  we  rely  on  the  Federal  Reserve  Board's 
series  of  debits  to  individual  accounts  in  141  cities  outside  of  New  York 
City.  It  includes  several  items  which  should  for  our  purpose  be  excluded 
and  is  otherwise  not  quite  what  we  want.  In  particular,  it  is  neither  free 
from  the  influence  of  speculative  transactions  nor  strictly  comparable 
with  the  series  with  which  we  are,  nevertheless,  about  to  relate  it.4 
But  still  we  may  put  our  trust  in  it  with  a  lighter  heart  than  in  the  clearing 
figures  previously  used  (see  Chart  XL VI). 

The  first  thing  that  strikes  us  is  the  weakness  of  the  "normalizing 
reaction"  to  the  war  disturbance.  In  fact,  there  is  hardly  any.  The 
annual  figure  for  1921  is  about  25  per  cent  below  the  obviously  abnormal 

1  See  International  Abstract  of  Economic  Statistics,  Conference  of  Economic  Services 
(London,  1934)  or  Special  Memorandum  83  of  the  London  and  Cambridge  Economic 
Service. 

2  See  H.  Neisser,  op.  cit.t   (Welturirtschaftliches  Archiv,   1920).     Before  August  1926 
that  was  not  so  in  the  case  of  the  dollar,  because  the  Reichsbank  rigidly  held  it  at  4.20. 

8  Bank  rates  and  other  points  will  come  up  for  discussion  later  on;  see  below,  sub  III. 
4  This  is  obvious  in  the  case  of  the  national  income,  but  it  should  be  emphasized  that  it 
is  also  true  of  the  deposit  series. 


818 


BUSINESS  CYCLES 


one  for  1920,  but  this  fall,  unlike  that  in  price  level  or  interest  rates,  did 
little  more  than  eliminate  the  effect  of  the  postwar  boom  and  the  1919 
figure  was  surpassed  already  in  1923.  No  other  series  shows  with  equal 
force  the  absence  of  any  significant  postwar  "deflation"  or  the  fact 
that  the  credit  apparatus  was  by  the  war  expanded  for  good,  which 


NET  DEM  A  NODE  POSITS 

I     I     I     I 

ANNUAL  AGGREGATE  REALIZED  INCOME 


HOURLY  EARNINGS  IN  MANUFACTURING 


1919    1920    1921    1922    1923    1924   1925    1926    1927   1928    1929    1930    1931    1932    1933    1934  1935 
CHART  XLVI.— United  States  (see  Appendix,  p.  1069). 

stands  out  particularly  impressively  if  we  compare  with  it  the  behavior 
of  the  English  series  of  country  plus  provincial  clearings  (see  Chart  LI). 
If,  then,  we  accept  this  as  a  datum,  the  rest,  viz.,  an  ascent  from  a  monthly 
average  of  16.6  billions  in  1922  to  a  monthly  average  of  27.7  billions  in 
1929,  substantially  steeper  than  that  of  outside  bank  clearings  from  1900 
to  1913,  also  testifies  to  the  unexhausted  powers  of  that  apparatus, 
though  a  more  modest  rate  of  increase  would  have  been  expected  by  us. 
A  straight-line  descriptive  trend  would  tolerably  fit  the  logs  of  the  figures. 
The  Kitchins  show  well  and  the  rise  of  the  fourth  Juglar  (1925)  is  indi- 
cated by  what  but  for  1928  and  19291  would  look  like  a  new  level. 

1  The  figures  of  these  years  are  of  course  the  ones  most  influenced  by  speculative 
transactions. 


1919-1929  819 

Our  chart  uses  the  Copeland-Crum  figures  for  realized  minus  imputed 
national  income  (see  Appendix),  but  no  difference  would  be  made  in  our 
rough  conclusions  if  those  of  Professor  Kuznets  (op.  cit.,  p.  8,  Col.  3  of 
Table  I)  had  been  used.  This  series  also  testifies  to  the  permanence  of 
the  monetary  revolution  that  had  occurred  but  it  presents  a  reduced  and, 
since  this  is  primarily  due  to  the  exclusion  from  it  of  capital  gains,  a  truer 
picture  of  it.1  The  postwar  boom  which  affected  the  figures  for  1919 
and  1920 — the  latter,  69  billions,  was  about  12  per  cent  above  the  former 
— left  national  income  for  1921  at  exactly  the  level  of  1918  (57.4). 
Then  follows  an  unbroken  ascent,  from  61.2  billions  in  1922  to  87.6  in 
1929.  The  short-run  behavior  corresponds — account  being  taken  of 
the  statistical  differences  between  the  two  series — to  that  of  debits  and 
would  do  so  still  more  if  undivided  profits  were  included. 

It  might  seem  that  the  pay-roll  index  charted  displays  a  still  more 
moderate  increase  and  thus  raises  an  additional  question.  But  on  closer 
inspection  it  will  be  seen  that  this  is  not  so.  Both  national  income  and 
(factory)  pay  rolls  fell  1920-1921  more  strongly  than  they  rose  in  1919- 
1920  (unlike  debits).  The  percentage  rise  and  fall  were  both  more 
pronounced  in  the  latter  than  in  the  former  case.  But  in  1922  pay  rolls 
also  recovered  at  a  much  greater  rate  than  did  national  income,  so  much 
so  that  not  merely  the  dip  in  1924,  which  occurs  only  in  the  pay-rolls 
index,  but  also  its  behavior  for  the  rest  of  the  time  explains  itself  naturally 
in  the  light  of  a  reaction  to  that  increase :  surveying  as  a  whole  the  stretch 
between  (for  pay  rolls:  middle  of)  1921  and  (for  pay  rolls:  third  quarter 
of)  1929,  we  find  substantial  parallelism.  Nevertheless,  there  possibly 
was,  apart  from  the  influence  of  that  difference  in  time  shape  and  from 
year-to-year  variations,2  also  a  significant  change  in  the  longer  run  rela- 
tion between  pay  rolls  and  income  which  reveals  itself  in  a  comparison 
made  by  Professor  Copeland  of  pay  rolls  in  banking  and  in  nonfarm 

1  In  the  long  run,  or  by  its  "underlying  tendency"  or  its  trend  values,  national  income 
in  terms  of  current  units  of  currency  may  be  said  to  measure  that  component  of  prices  and 
values  which  is,  at  least  immediately,  attributable  to  monetary  causes,  see  Chaps.  VIII 
and  X. 

2  Professor  M.  A.  Copeland,  National  Wealth  and  Income,  Journal  of  the  American 
Statistical  Association  for  June  1935,  pp.  385-386,  notices  that  the  ratio  of  pay  roll  to 
realized  income  and  the  proportion  of  total  income  received  by  the  poorest  90  per  cent 
of  the  nonfarm  population  (but  gross  farm  income  moved  much  as  income  of  industrial 
workers,  see  M.  Ezekiel,  op.  tit.,  chart  on  p.  140)  show  similar  increments  and  adds  that 
from  1918  to  1929  the  latter  rises  and  falls  with  the  ups  and  downs  of  business.     This  is  so, 
but  it  must  be  entirely  due,  first,  to  the  inclusion  of  the  abnormal  year  1919,  second,  to 
the  abnormally  strong  increase  in  wages  in  1923,  and,  third,  to  the  exclusion  of  undivided 
or  unwithdrawn  profits.     Hence,  the  writer  cannot  follow  Professor  Copeland  in  consider- 
ing that  finding  relevant  to  "the  Hobsonian  view  [now  unfortunately  sponsored  by,  among 
lesser  lights,  Dr.  Ezekiel,  J.  A.  S.]  that  increased  concentration  in  periods  of  prosperity  is 
responsible  for  a  disproportionate  volume  of  saved  income,"  however  much  he  rejoices  on 


820  BUSINESS  CYCLES 

industry  with  the  comparable  part  of  total  realized  income.1  This  rela- 
tion was,  at  about  73  per  cent,  in  the  long  run  fairly  stable  before  the 
war.  We  should  expect  and  actually  find  that  war  expenditure  dis- 
turbed, i.e.,  at  first,  decreased  it — we  remember  that  adventitious,  in 
contrast  to  entrepreneurial,  demand  always  tends  to  do  that.  But  reac- 
tion against  this  is  observable  in  1917,  and  the  improvement  in  the 
bargaining  position  of  labor,  the  labor  policies  of  the  various  war  boards 
and  the  conditions  of  the  postwar  boom  carried  it  much  beyond  that 
prewar  level  to,  if  we  may  trust  the  figures,  about  84  per  eent.  Some 
decline,  though  owing  to  the  change  in  Kondratieff  phase  not  necessarily 
to  prewar  levels,  was  due  within  the  process  of  general  normalization 
and  may  in  fact  be  seen  in  those  figures,  although  the  ratio  remained 
much  above  prewar  normal  in  1929  (77.9).  To  this  and  cognate  subjects 
we  shall  return  below. 

b.  Corporate  accumulations  are,  as  stated  above,  not  included  in  the 
income  figures  discussed.2  We  may  use  their  variations  as  indicators 
of  the  variations  in  all  accumulations.3  In  doing  so  we  must  of  course 
bear  in  mind  that,  as  presented  in  the  official  Statistics  of  Income, 


general  grounds  in  Professor  Copeland's  refusal  to  support  the  implications  of  that  unsound 
doctrine.  But  the  theorem  that  during  recession  aggregate  profits  converge  toward  zero 
and  during  depression  tend  to  become  negative  is  not  so  very  far  removed  from  reality. 

1  Professor  Copeland's  series  and  the  argument  that  follows  in  the  text  involve  the 
use  of  two  different  and  imperfectly  comparable  series,  Professor  King's  from  1909  to  1920 
and  Professor  Copeland's  own  from  1920  to  1929.  If  we  went  on,  a  third,  Professor  Kuz- 
nets',  would  have  to  be  used. 

a  That  is  to  say,  they  have  not  been  added.  But  as  far  as  they  are  not  used  for  the 
purpose  of  increasing  cash  items  (and  as  far  as  they  really  exist),  they  of  course  eventually 
reappear  in  individual  incomes,  although  some  authors  argue  as  if  they  were  stored  up. 

3  No  estimate  will  be  offered  of  household  savings  but  only  certain  indications,  because 
we  simply  do  not  know  enough  about  them,  particularly  as  regards  the  roughly  89,000 
(1929)  individual  net  incomes  above  $50,000.  The  most  ambitious  attempt  that  has  been 
made  (Levin,  Moulton,  and  Warburton,  America's  Capacity  to  Consume,  1934,  followed 
by  two  interpretative  volumes  by  H.  G.  Moulton,  Formation  of  Capital,  1935,  and  Income 
and  Economic  Progress,  1935)  has,  statistically  and  theoretically,  been  so  severely  criticized 
that  it  is  unnecessary  to  explain  why  we  do  not  avail  ourselves  of  the  results.  In  particular, 
the  author  has  little,  if  anything,  to  add  to  Mr.  H.  H.  Villard's  article  on  Dr.  Moulton's 
Estimates  of  Saving  and  Investment,  American  Economic  Review  for  September  1937,  which 
completely  disposes  of  the  contention  that  there  was  any  excess  of  net  monetary  savings 
over  net  productive  investment,  and  thus  of  one  type  of  oversaving  theories  as  applied  to 
the  postwar  period.  Concerning  another  type,  it  is  relevant  to  note  that,  as  one  of  the 
co-authors  of  the  first  of  the  volumes  mentioned,  Mr.  C.  Warburton,  has  pointed  out 
(Capacity  to  Consume,  p.  Ill,  and  again  in  Trend  of  Savings,  1900-1929,  Journal  of 
Political  Economy,  1935,  p.  84  et  seq.),  there  was  no  significant  "trend"  in  the  percentage  of 
savings  for  the  whole  period  1917  to  1929,  even  in  the  data  of  the  Brookings  investigation, 
if  it  were  admissible  to  speak  of  trends  in  the  case  of  a  series  extending  over  13  years.  There 
would  be  a  downward  one,  or,  for  1922  to  1929,  a  very  slightly  rising  one,  if,  as  we  must, 


1919-1929  821 

they — net  incomes  minus  cash  dividends  paid  to  the  individuals — are  in 
some  cases  the  product  of  arbitrary  decisions  and  in  others  of  irrational 
bookkeeping  routine.  Obsolescence  in  particular  can  hardly  ever  be 
adequately  taken  account  of  either  by  the  executives  with  whom  those 
decisions  rest  or  by  the  observer.  Moreover,  the  undistributed  part  of 
net  income  not  only  fills  the  function  of  accumulation  in  our  sense  but 
also  that  of  an  equalization  fund,  so  that  they  really  ought  to  be  referred 
not  to  a  year  but  to,  say,  a  Juglar.  Finally,  the  figures  must  be  cor- 
rected for  the  substantial  difference  that  exists  between  the  usual  depre- 
ciation at  cost  and  the  appropriate  depreciation  at  current  prices.  This 
has  been  done  by  Mr.  S.  Fabricant,  whose  corrected  series  for  all  corpora- 
tions except  tax-exempt  and  life  insurance  companies  we  are  going  to  use.1 
The  1919  boom  carried  them  to  a  peak  that  was  never  reached  again — 
$3,310  millions,  a  figure  which  illustrates  the  equalization  function  of 
undivided  surplus.  Whether  or  not  corporate  business  intended  to  pre- 
pare itself  to  meet  the  difficulties  that  were  to  come  by  accumulating 
that  "reserve,"  the  latter  was  actually  almost  wiped  out  in  1920  (—-$10 
millions)  and  1921  (  —  $3,240  millions).  Starting  with  a  practically  clean 
slate  in  1922,  we  have,  for  that  and  the  two  following  years,  which  about 
completed  the  Juglar  recovery,  a  total  of  $4,090  millions.  The  next 
3  years,  roughly  covering  the  prosperity  phase  of  the  fourth  Juglar, 
added  $4,760  millions,  over  half  of  them  in  1925.  This  accords  well  with 
expectation.  But  the  contribution  of  1928,  $2,040  millions  and,  though 
less  so,  that  of  1929,  $970  millions,  are  above  expectation2  and  must  be 
included  in  our  growing  list  of  abnormalities  which  showed  during  those 

we  exclude  capital  gains.  (See  Chart  II  on  p.  101,  inference  from  which  involves,  however, 
the  hypothesis  that  the  percentage  of  national  income  saved  is  roughly  equal  to  the  per- 
centage of  income  derived  from  property.)  "With  capital  gains  excluded,  this  percentage 
[of  savings]  was  lower  throughout  the  decade  1920-1929  than  during  the  prewar  years 
1909-1914"  (p.  100).  This  is  in  itself  sufficient  to  destroy  the  case  for  either  an  over- 
saving or  an  overinvestment  inference  from  this  material.  We  may  take  the  opportunity 
to  refer  to  Mr.  W.  H.  Lough's  critique  of  the  Brookings  estimates  of  household  savings  in 
his  most  helpful  book  on  High-level  Consumption,  1935  (Appendix  G)  and  to  add  that  his 
own  estimates  (see  below),  although  owing  to  the  use  of  a  different  concept  of  saving  and 
for  other  reasons,  far  above  what  savings — which,  be  it  recalled,  exclude  among  other 
things  sums  assembled  for  residential  construction  for  owner  occupancy — in  our  sense  can 
have  amounted  to,  yet  fail  to  display  an  upward  trend  for  1919  to  1929  (though  they  of 
course  increase  from  1922  to  1928)  and  that  their  percentage  share  in  his  grand  totals  of 
realized  income  displays  a  falling  one. 

1  Measures  of  Capital  Consumption,   1919  to  1933,  National  Bureau  of  Economic 
Research,  Bulletin  60,  for  June  30,  1936,  p.  12.     The  limitations  of  the  material  and  the 
difficulties  of  handling  it  are  fully  discussed  in  that  study,  to  which  the  reader  is  hereby 
referred. 

2  They  are,  however,  in  accord  with  the  Kitchin  phases,  which  in  this  series  show  very 
well  throughout.    The  dips  of  1924  and  1927  are  strongly  marked. 


822  BUSINESS  CYCLES 

two  years.  This  makes,  for  the  eight  years,  $1,485  millions  per  year  or, 
for  the  eleven  years,  $1,084  millions. 

There  is  no  point  in  stressing  either  figure,  because  neither  1922  to 
1929  nor  1919  to  1929  forms  any  unit  that  has  any  cyclical  meaning. 
But  there  is  point  in  stressing  their  comparative  smallness:  the  first 
must,  moreover,  be  judged  in  the  light  of  the  fact  that  the  span  to  which 
it  applies  does  not  contain  the  Juglar  depression,  which  in  this  case,  as 
we  shall  see,  more  than  wiped  out  the  total.  Although  we  cannot  in  so 
short  a  series  speak  of  a  trend  in  our  sense — a  result  trend — but  only  of 
Juglar  and  Kitchin  fluctuations,  it  is  worth  while  also  to  note  that  a 
formal  trend  through  1922  to  1928  would  not  display  a  significant  inclina- 
tion.1 And,  if,  from  the  standpoint  of  oversaving  theories,  the  figure  of 
1928  were  fastened  upon  in  connection  with  an  explanation  of  subsequent 
vicissitudes,  we  should  ask  in  reply  whether  a  figure  only  about  half  a 
billion  above  the  average  could,  even  from  that  standpoint,  be  looked 
upon  as  adequate,  and  why  the  still  higher  figure  of  1925  did  not  prevent 
a  perfectly  normal  Juglar  prosperity  and  a  supernormally  active  Juglar 
recession.  The  order  of  magnitude  of  the  total  precludes  any  retreat 
on  cumulative  effects.  It  was  not  even  enough  to  prevent  a  substantial 
increase  in  long-term  debt,  in  spite  of  the  booming  stock  markets. 

Comparing  the  course  of  corporate  accumulations  which  has  been 
just  described  with  the  evidence  embodied  in  Chart  XLIII,  we  see  what 
has  often  been  observed,  viz.,  that  there  is  fair  correspondence — though 
opinions  may  differ  about  the  nature  and  average  amount  of  the  lag — 
between  the  variations  in  those  accumulations  and  in  the  production 
of  durable  goods,  as  there  also  is,  let  us  add,  between  the  former  and 
expenditure  on  plant  and  equipment.  This  is  of  course  as  it  should  be, 
but  the  inference  that  might  be  drawn  is  much  weakened  by  the  fact  that 
covariation  with  total  production  of  manufacturing  industry  is  just  about 
equally  good  and  covariation  with  consumers'  goods'  production  almost 
so.  The  three — and  many  other — quantities  simply  move  together 
within  one  process,  as,  to  use  Professor  Leontief  's  happy  phrase,  soldiers 
do  in  a  marching  battalion,  and  there  is  little  justification  for  picking  out 
the  relation  between  any  two  of  them  and  still  less  for  interpreting  it 
causally.2 

1  Cash  dividends  also  display  the  Juglar  and  Kitchin  phases,  but  only  slightly :  they 
increase  from  the  last  quarter  of  1921  to  the  end  of  1929  in  an  almost  straight-line  manner. 
They  were  greater  by  about  3.1  billions  in  1929  than  in  1922,  thus  absorbing  most  of  the 
net  increase  in  the  net  revenue,  exclusive  of  taxes,  over  the  period.     Corporate  accumula- 
tion failed  to  keep  step  with  corporate  distribution.     The  implications  of  this  must,  no  doubt, 
be  qualified  by  taking  account  of  the  household  savings  of  shareholders.     But  what  we  know 
about  them  does  not  invalidate  the  obvious  conclusion,  viz.,  that,  so  far  as  this  source  of 
income  goes,  the  percentage  "saved"  decreased  as  the  income  itself  increased. 

2  But  we  shall  not  be  surprised  to  find  that  the  more  usual  "causal"  relation  has  now 


1919-1929  823 

c.  It  is  of  some  interest  finally  to  try  to  form  an  idea,  although  it 
can  only  be  a  very  imperfect  one,  about  the  relation  between  total  realized 
income  (which  excludes  corporate  accumulations)  and  consumers' 
expenditure,  which  will  also  shed  some  light  on  the  order  of  magnitude 
of,  and  on  the  variations  in,  net  savings  of  all  households  combined. 
An  indication  is  afforded  by  the  fact  that  the  percentage  change  in 
department-store  sales  was  from  about  the  middle  of  1923  to  the  first 
months  of  1929  practically  equal  to  the  percentage  change  in  nonagri- 
cultural  income.1  Department-store  sales  rose  more  and  fell  less  than 
nonagricultural  income  in  1920  and  1921,  but  then  the  indices  of  both 
(1928-1929  =  100)  got  and  kept  together  until  the  second  quarter  of 
1929 — after  which  they  fell  so  closely  in  step  that  the  curves  in  1932 
practically  coincide — when  nonagricultural  income  rose  slightly  more 
than  sales.  There  is  no  systematic  tendency  for  variations  in  sales  to 
lag  behind  or  to  become  relatively  smaller  as  aggregate  income  increases. 
The  suggestion  implied  is  substantially  confirmed,  if  due  account  be 
taken  of  consumers'  outlay  on  those  durable  goods  for  which  department- 
store  sales  are  not  typical — the  main  instances  being  outlay  on  passenger 
cars,  which  is,  and  acquisition  of  homes,  which  is  not,  as  a  rule  included 
in  the  estimates  of  consumers'  expenditure.  This  is  done  by  the  figures 
compiled  by  Lough,  King,  Kuznets,  and  Warburton.2  Moreover,  those 
figures  indicate,  although  the  fact  is  veiled  by  differences  in  definitions 
and  classifications,  not  only  that  the  absolute  values  of  aggregate  realized 
income  and  aggregate  expenditure  on  consumers'  goods  displayed  no 

been  turned  around  by  some  theorists  of  consumers'  credit  so  as  to  read  that  stimulation 
of  consumption  will  stimulate  investment  and  even  "  saving"  and  is,  in  fact,  the  life-giver 
of  the  economic  process.  There  is  as  much  truth  as  there  is  error  in  any  such  statements, 
the  very  inadequacies  of  which  then  produce  "discoveries"  that  are  pregnant  with 
policies. 

1  This  finding  is  due  to  Dr.  Louis  Bean.     See  Nonagricultural  Income  as  a  Measure  of 
Domestic  Demand,  by  Bean,   Bollinger,  and  Wells,  U.  S.  Department  of  Agriculture, 
Agricultural    Adjustment    Administration,    Agricultural    Industrial    Relations    Section, 
June  1937,  p.  8.     The  index  of  nonagricultural  income  is  not  strictly  comparable  with 
our  realized  income  but  the  contours  are  the  same. 

2  W.  II.  Lough,  op.  cit.,  p.  26,  chart  on  p.  27  (Total  Outgo,  roughly  equal  to  his  realized 
income,  which  includes  imputed  rentals,  to  be  compared  with  "commodities"  plus  "intangi- 
bles"); W.  J.  King,  National  Income  and  Its  Purchasing  Power,  1930;  S.  Kuznets,  National 
Income  (1938),  Table  15,  p.  53:  consumers'  outlay  (in  his  sense)  fluctuates  1921  to  1929 
around  88  per  cent,  the  last  year  displaying  the  larg  st  percentage  (90.7),  and  after  rear- 
rangement to  fit  our  concepts  around  95  per  cent  of  his  national  income.     C.  Warburton, 
How  the  National  Income  was  Spent,   1919-1929   (Journal  of  the  American  Statistical 
Association,  March  1935,  Papers  and  Proceedings  of  the  96th  meeting,  p.  177,  incomparable 
with  the  other  estimates  as  they  are  between  each  other)  obtains  for  consumption  in  per 
cent  of  his  national  product  a  figure  varying  between  79.8  (1919)  and  102.6  (1921).     Since 
1925  shows  85,  this  is  the  only  series  in  which  the  percentage  is  markedly  higher  in  depres- 
sion and  recession  than  in  prosperity. 


824  BUSINESS  CYCLES 

tendency  to  drift  apart,  as  some  theories  postulate  they  should,  but  also 
that  they  did  not  differ  so  very  much  from  each  other.1 

The  reader  will  be  inclined  to  attribute  these  results  to  statistical 
miscarriages  and  wonder  how  either  of  them  can  be  true,  considering, 
on  the  one  hand,  that  we  know  from  common  experience  that  households 
do  save2 — and,  so  we  are  taught,  out  of  increasing  incomes  more  than 
out  of  decreasing  ones — and  on  the  other  hand,  that  rearrangements  in 
the  timing  of  consumers'  expenditure — such  as  is  involved  in  a%  shift  of 

1  Mr.  Lough  (op.  cit.,  see  in  particular  the  basic  tabulation  in  Appendix  A  and  p.  806) 
arrives,  it  is  true,  for  1919  to  1929  at  a  total  of  "  saving  from  realized  income"  which  is 
not  far  short  of  100  billions.     Relatively  minor  points  apart,  this  is,  of  course,  due  to  his 
treatment  of  increases  in  cash  holdings,  of  all  life  insurance  premiums  and  of  payments 
for  the  acquisition  of  real  estate,  and  if  the  total  is  scaled  down  to  fit  our  concepts,  the 
difference  between  income  and  expenditure  reduces  to  a  quantity  of  the  order  of  magnitude 
of  about  5  per  cent  of  the  former.     We  may  illustrate  the  point  also  by  a  rearrangement  of 
Mr.  Warburton's  figures,  op.  cit.,  p.  178.     If  we  deduct  from  his  totals  for  consumers' 
goods  and  services  the  item  of  imputed  rentals  and  if  we  add  to  them  the  item  of  residential 
construction  (not  all  of  it  ought  to  be  added,  of  course,  but,  on  the  other  hand,  we  neglect 
other  items  which  should)  we  find  that  the  result  comes  fairly  close  to  the  Copeland-Crum 
revised  series  of  realized  income.     The  differences  (income  —  consumers  outlay)  then  are  in 
billions:  for  1919,  1.8;  for  1921,  minus  3.4;  for  1923,  minus  1.8;  for  1925,  minus  4;  for  1927, 
0.8;  for  1929,  3.2.     This  may  read  absurd  (especially  the  minus  figures  do)  but  is  not  in 
the  least.     That  realized  income  should  fall  short  of  consumers'  expenditure  is  not  at  all 
unlikely  in  a  period  in  which  everyone  rushed  into  debt  and,  incidentally,  impugned  over- 
saving.    And  there  is  (though  plenty  of  roughness)  no  statistical  mistake  in  the  comparison 
of  two  substantially  independent  and  comparable  aggregates,  nor  does  it  follow  ipso  facto 
(this  would  mean  begging  the  question)  that  the  Copeland-Crum  estimates  are  too  low. 
No  inference  can,  of  course,  be  drawn  from  the  detail  of  the  behavior  of  those  differences. 
They  do  show,  however,  that  in  the  present  state  of  our  information  presence  or  absence 
of  net  savings  is  a  matter  of  margins  of  error  and  this  illustrates  the  point  we  wish  to  make 
quite  sufficiently.     Into  the  shifts,  so  interesting  from  many  standpoints,  which  we  observe 
between  different  classes  of  consumers'  expenditure,  we  cannot  and  need  not  go.      The 
one  that  is  relevant  to  our  purpose,  the  shift  toward  durables,  has  been  mentioned  and  is 
in  any  case  obvious. 

2  But  since  earmarkings  for  the  rainy  day  and  for  the  acquisition  of  durable  goods, 
homes  included,  and  nonspending  of  capital  gains  all  do  not  come  within  our  concept  of 
savings — the  writer  trusts  that  the  reader  remembers  the  reasons  we  have  for  this — the 
fact  is,  for  the  period  under  discussion,  not  quite  so  palpable  as  it  seems.     The  first  two 
items  will  cover  much  of  what  is  usually  considered  as  the  savings  of  the  lower  and  middle 
income  groups;  the  third  will  cover  much  of  the  so-called  savings  of  the  middle  and  higher 
groups.     We  are  speaking  of  this  country  and  a  period  of  an  all  pervading  speculative 
attitude  and  of  uncritical  optimism.     The  successful  lawyer,  doctor,  and  business  execu- 
tive, all  speculated  on  the  stock  exchange  and,  as  long  as  things  went  well,  were,  say,  up 
to  an  income  of  $50,000  exclusive  of  such  gains,  under  great  temptation  to  look  upon  their 
current  earnings  as  a  fund  for  current  expenditure.    This  may  have  been  different  in  the 
case  of  seasoned  property  incomes  of  the  larger  and  largest  sizes.     In  1929  there  were  513 
individual  income  tax  returns  stating  incomes  of  a  million  and  over,  but  only  75  in  1924. 
If  we  look  upon  the  latter  incomes  as  the  "seasoned"  ones — assuming  that  recipients  of 
war  profits  had  also  been  sobered  by  that  time — we  may  guess  that  the  prevalent  idea 


1919-1929  825 

consumers'  demand  toward  more  durable  goods — underspending  in  deep 
depression,  increasing  replacement  of  saving  by  insurance,  and  direct 
taxes  paid  out  of  incomes1  must  all  of  them  create,  if  not  savings  or  not 
savings  in  our  sense,  at  least  discrepancies  between  our  aggregates,  and 
raise  the  figures  of  realized  income — increasingly  even — above  those  of 
consumers'  expenditure.  The  answer  is  simple,  however.  All  those 
items  were  to  a  large  extent  compensated  by  dissaving  (mainly  the 
spending  of  capital  gains)  and  borrowing,  the  latter  including  some  that 
professed  to  be  for  business  purposes.  There  obviously  were  in  all 
strata  of  society  very  many  people  who  "lived  above  their  means/'2 
as  is  shown  by  the  ready  response  to  any  boom  or  slump,  particularly 
in  stock  markets,  of  the  sales  of  industries  catering  for  "luxury"  demand.3 
We  thus  return  with  added  emphasis  to  the  opinion  already  submitted 
for  the  prewar  time  in  an  earlier  chapter,  viz.,  that  the  amount  of  net 
households'  savings,  as  distinguished  from  firms'  accumulation,  is  being 
greatly  exaggerated  even  by  the  more  sober  ones  of  the  current  estimates. 
In  support,  we  may  point  again  to  the  evidence  there  is — as,  to  a  lesser 
extent,  there  was  for  the  prewar  time — for  households'  straining  their 
resources  in  order  to  expand  their  consumption. 

about  the  size  of  rich  men's  genuine  savings  applies  fairly  well  to  these,  although  some 
households  known  to  be  in  that  category  spent  pretty  freely  on  their  own  consumption, 
— the  higher  the  income,  the  less  its  relative  purchasing  power  as  compared  with  Europe — 
and  others  on  other  people's.  But  those  ideas  hardly  apply  to  the  rest.  The  drift  of  the 
argument  would,  of  course,  not  be  affected  if  we  had  chosen  a  lower  than  the  million  limit. 

1  Of  underspending  there  is  some  evidence  for  1921.     But  otherwise  there  was  np  year 
of  deep  depression  and,  hence,  there  cannot  have  been  significant  underspending.     Rear- 
rangements of  the  time  shape  of  expenditure  were  mostly  effected  by  borrowing,  but  they 
must  have  had  some  influence  in  causing  temporary  withholdings.     The  earmarking  of 
sums  for  payments  of  life  insurance  premiums  is,  of  course,  saving  in  some,  but  not  in  other, 
cases.     State  and  Federal  taxes  are  (except  for  1919)  more  than  balanced  by  the  "public 
consumers'  goods,"  public  buildings,  highways,  and  streets. 

2  That  turn  of  phrase  applies  strictly  to  part  only  of  the  cases  we  refer  to  and  becomes 
misleading  beyond  it.     Nobody,  for  example,  thought  he  was  living  above  means,   still 
less  that  he  was  dissaving,  if  he  bought  himself  a  home  from  realized  appreciation  of  stocks. 

3  This  is  true  for  the  "luxuries"  of  all  income  brackets,  of  course.     It  is,  however,  not 
less  but  more  true  for  those  of  the  higher  ones.     It  was  not  the  third-class  but  the  first- 
class  hotels,  for  example,  which  were  overcrowded  in  1929  and  in  which  a  visitor  had  a  whole 
floor  to  himself  in  1931  to  1938.     Even  obviously  temporary  gains  are  readily  spent  and 
play  a  large  role  in  the  financing  of  the  consumption  precisely  of  the  higher  income  levels. 
That  the  lower  and  lower  middle  classes  spent  a  larger  percentage  of  their  income    on 
consumption  than  before  the  war  may  perhaps  be  inferred  from  the  expansion  in  the 
production  of  their  luxury  goods  and  is  in  itself  plausible,  owing  to  the  growth  of  social 
insurance  and  to  an  unmistakable  change  in  attitude.     But  the  former  may  also  be  due 
to  a  shift  in  taste  and  such  indications  as  we  have  are  hardly  trustworthy.     See,  for  exam- 
ple, Mr.  Warburton's  argument  in  favor  of  this  view,  Trend  of  Saving,  p.  97.     Nothing 
can  be  concluded  from  figures  such  as  are  there  presented  without  committing  a  contempt 
of  court  in  matters  of  statistical  principle. 


826  BUSINESS  CYCLES 

According  to  Mr.  Lough,  short-term  consumers'  debts  increased  by 
about  4  billions  from  1919  to  1929.1  This  includes  loans  against  life 
insurance  policies  and  federal  loans  to  veterans  as  well  as  installment 
paper,  open  accounts,  overdue  items,  loans  by  loan  agencies,  pawn- 
brokers, personal  loans  by  commercial  banks,  but  not  loans  against 
building  and  loan  shares  or  home  mortgage  loans.  Very  much  higher 
estimates  have  been  published,2  but  the  one  quoted  suffices  to  indicate 
a  tendency  which  becomes  still  more  significant  by  virtue  of  the  fact  that 
it  is  the  years  of  rapidly  rising  incomes  which  display  the  largest  increases : 
1923  (776  millions:  maximum  increase),  1928  (609),  and  1929  (691). 
The  most  obvious  single  symptom  is  the  growth  of  buying  on  install- 
ments, which  spread  from  motorcars  and  the  new  household  goods — 
refrigerators  and  so  on — to  clothing  and,  while  it  was  spreading,  involved 
anticipation  of  future  income.  Exact  figures  exist  only  for  individual 
industries.  For  the  total  we  have  estimates  which  vary  widely :  the  first 
that  was  based  on  a  careful  investigation — that  of  M.  V.  Ay  res — was 
5.7  billions  for  1925.  It  has  been  criticized,  and  reduced  to  4.875  by 
Professor  Seligman,  still  more  so  by  later  writers.3  We  know,  however,4 
that  in  1927,  23,779  retail  stores  whose  total  sales  were  over  4.7  billions 
and  which  included  a  number  of  "cash  and  carry"  establishments,  sold 
9.2  per  cent  of  the  total  on  installments,  32.2  per  cent  on  open  credit 
arrangements,  and  only  58.6  per  cent  for  cash.  Motorcar  dealers  sold 
nearly  50  per  cent,  or  according  to  another  estimate  585  per  cent  on  install- 
ment; furniture  dealers,  57.7;  and  in  the  lumber  and  building  materials 
trade  open  credit  was  given  for  90  per  cent  of  the  sales. 

The  role  of  both  borrowings  and  dissavings  in  the  processes  of  the 
twenties  and  their  relation  to  the  subsequent  breakdown  is  obvious. 
But  considering  the  prevalence  of  the  view  that  consumers  displayed 
want  of  alacrity  in  responding  to  increase  in  monetary  incomes  or,  for 
that  matter,  in  business  revenue  and  that  this  was  a  source  of  troubles, 

1  See  op.  cit.,  p.  312.     The  increase  is  from  5.4  billions  to  9.4,  part  of  it  being,  of  course, 
due  to  carrying  charges. 

2  The  highest  that  has  come  to  the  notice  of  the  writer  is  that  of  F.  W.  Ryan  (Internal 
Debts  of  the  United  States)  which  for  1929  is  22  billions. 

8  These  figures  refer  to  total  volume  of  installment  sales,  not  to  installment  paper  out- 
standing. Every  series  of  estimates  the  writer  has  seen  displays  a  rising  "trend"  from 
1919  to  1929,  Mr.  Lough's  to  about  double.  The  significance  of  this  must  not  be  exag- 
gerated. A  new  method  of  paying  for  durable  goods  was  simply  gaining  ground.  But 
while  it  expanded,  it  meant  buying  beyond  the  funds  available  from  consumers'  income 
streams,  and  this  is  enough  for  us.  No  need  to  say  that  bulges  occur  in  the  relatively 
prosperous  years  in  which  people  should,  according  to  prevailing  theories,  be  busily  saving. 

4  National  Retail  Credit  Survey,  Part  III,  1930,  Department  of  Commerce. 

6  C.  C.  Hanch,  Composite  Experience  of  Automobile  Finance  Companies,  1927.  It 
had,  according  to  the  same  source,  been  75.5  in  1925.  The  rapid  growth  of  consumers' 
finance  companies  is  in  itself  a  significant  fact. 


1919-1929  827 

the  facts  glanced  at,  fragmentary  as  they  are,  have  an  importance  in 
themselves,  irrespective  of  all  theory.  Though  we  do  not  hold,  of  course, 
that  any  dire  consequences  would  have  followed  from  a  higher  amount 
or  from  an  increasing  percentage  rate  of  savings,  we  may  still  note  that 
oversaving  theories  would  not  apply,  even  if  they  were  logically  unexcep- 
tionable. 

d.  In  Germany,  the  estimate  of  national  income  which  holds  the  field 
is  that  of  the  federal  statistical  office  (Reichsamt) . l  Since  the  first 
figure  is  for  1925  (59  billion  marks),  we  have  five  in  all.  There  was  strong 
increase  to  1928  (73.4)— and  a  small  one  to  1929  (73.6)— which  would  be 
somewhat  accentuated  by  the  addition  of  reparation  payments  which 
have  been  excluded.  The  national  wage  bill  (which  includes  all  salaries 
and  is  hence  affected  by  the  increase,  relative  as  well  as  absolute,  of 
salaried  employees  of  all  ranks)  also  increased  steadily  from  33.7  billions 
to  42.6  in  1928  (and  43  in  1929)  hence  by  about  2  per  cent  more  than 
national  income  or  (if  that  increasing  weight  of  salaries  be  taken  into 
account)  roughly  as  the  latter.2  This  too  is  perfectly  normal,  although 
the  movements  from  year  to  year  reveal  various  abnormalities,  the 
most  important  of  which  is  that  the  increase  in  wage  bill  from  1927  to 
1928  was,  in  absolute  amount,  almost  equal  to  the  increase  in  national 
income.  In  spite  of  this,  however,  that  wage  bill  was  even  for  1928  not 
more  than  58  per  cent  of  national  income,3  which  considering  the  com- 
prehensiveness of  the  former  is  anything  but  high. 

In  comparing  the  national  income  with  consumers'  expenditure  we 
strike  the  same  phenomenon,  only  still  more  obviously,  which  we  observed 
in  the  American  case.  That  figure  of  national  income  includes  corporate 
accumulation.  A  careful  estimate  of  consumers*  outlay,4  which  excludes 

1  Monograph    (Einzelschrift),    No.    14.     Estimates   of  national  income  at  one  time 
threatened  to  become  a  political  affair,  and  it  was  possible  to  infer  from  the  figures  com- 
pilers presented  whether  they  were  siding  with  trade  unions  or  opposed  to  their  policies. 
The  one  we  use  seems  to  be  entitled  to  confidence.     Its  main  bases  are  income-tax  and 
wage-tax  statistics.     Double  counting  is  carefully  guarded  against.     This  does  not  mean, 
however,  that  it  is  strictly  comparable  to  the  American  series  though  variations  presumably 
are. 

2  The  figures  for  wages  are  not  quite  safe.     The  wage  tax  with  its  many  exemptions 
and  rebates  proves  on  investigation  a  less  valuable  ally  than  one  would  think.     Comparison 
with  the  prewar  time  when  it  did  not  exist  is  too  hazardous  a  venture  to  undertake,  since 
the  figure  of  wage  income  for  1918  is  highly  conjectural.     This  also  casts  doubts  on  the 
precise  significance  of  the  sharp  increase  from  1924  to  1925  (about  40  per  cent  in  the 
industrial  pay  roll). 

3  It  is,  however,  important  to  bear  in  mind  that  the  income  of  wage  and  salary  receivers 
was  considerably  increased  by  "social  rents"  and  pensions  which  are  not  included  in  the 
wage  bill  and  in  1928  amounted  to  11.2  per  cent  of  the  national  income,  a  largely  though 
not  wholly  additive  item. 

4  Walther  Lederer,  Was  verbrauchen  wir?  in  Die  Arbeit,  1932, 


8£8  BUSINESS  CYCLES 

residential  construction  for  owner  occupancy,1  nevertheless  comes  close 
to  it.     The  differences  (income  minus  outlay,  in  billions  of  marks)  for 

1925  to  1928  are:  1.3,  1.4,  1.9,  1.7.     If  taxes  paid  out  of  incomes  are 
taken  into  account  there  is  a  negative  figure  for  every  year.     This 
accords  with  an  investigation  of  the  Institut  fiir  Konjunkturforschung.2 
Now,  of  course,  it  will  be  held — it  has  been  held,  as  a  matter  of  fact — 
that  this  proves  only  that  the  income  figures  are  too  low.     They  may 
be.     But  this  cannot  be  inferred  from  that  fact  alone,  which  it  is  erroneous 
to  consider  as  logically  impossible  or  wildly  unlikely.     It  is  perfectly 
possible  and  fits  excellently  into  the  general  picture  of  the  situation. 
People  simply  borrowed — again  in  many  cases  ostensibly  for  productive 
purposes — and  dissaved. 

For  the  United  Kingdom  we  use  Mr.  Colin  Clark's  income  estimates 
for  1924  to  1929.3  We  start  (in  billions  of  pounds)  with  3.36,  then  have 
the  Juglar  jump  we  expect  in  1925  (3,7),  an  understandable  drop  in 

1926  (3.53),  which  was  roughly  made  up  for  in  1927  (3.67)  and  1928 
(3.64),   and  end  up  with  some  reflex  of  the   1929  spurt   (3.73).     The 
British  income-tax  system  had   (up  to  1927-1928,  see  note  below)   a 
smoothing  effect,  but  considering  this,  fluctuations  show  as  they  should. 
There  is  also  a  trace  of  the  rising  tendency  we  expect.     But  the  whole  is 
a  very  sober  affair  and  characteristically  different — so  much  is  clear,  in 
spite  of  the  lack  of  comparability  of  series — from  either  the  American 
or  the  German  case.     The  monetary  revolution  shows,   of  course,  by 
comparison  with  the  last  prewar  years.     But  no  machine  for  monetary 
expansion  had  been  set  up,  as' it  had  in  this  country,  which  would,  after 
liquidation  of  the  war,  go  on  by  its  own  momentum.     Pressure  on  the 
aggregate  income  by  the  monetary  factor  is,  on  the  contrary,  obvious — 
the  presence  of  a  component  that  worked  toward  continuity  with  prewar 
levels.     This  would  in  itself  suffice  to  justify  the  "prediction"  that  the 
subsequent  great  depression  should  be  comparatively  mild  and  short. 

1  Both  series  include  imputed  rentals. 

2  Vierteljahrshefte  zur  Konjunkturforschung,  vol.  V,  No.  4.     Mr.  Lederer  raises  various 
objections  to  the  result,  which  is  a  deficit  for  each  of  the  4  years.     These  objections  seem 
justified  in  part.     But  he  only  succeeds  in  turning  the  deficits  into  on  the  average  minute 
surpluses. 

8  See  National  Income  and  Outlay,  19S7,  p.  88.  This,  of  course,  implies  that  we  accept 
still  another  concept.  Mr.  Clark's  most  important  nostrum,  however,  the  addition  to 
national  income  of  public  revenue  from  indirect  taxation,  rates,  and  so  on,  we  will  excuse 
ourselves  from  accepting:  the  total  in  Table  35  has  been  subtracted  from  the  total  in  Table 
37  and  so  have  been  undivided  profits  (Table  85,  p.  187).  Incomparability  with  the 
American  or  the  German  series  need  not  be  stressed.  Discontinuance  in  1927-1928  of 
the  three-year-average  system  should  be  kept  in  mind.  Mr.  Clark  most  commendably 
attempts  to  pierce  the  veil  of  the  actual  tax  privileges  of  agriculture.  The  British  income 
tax  law  does  not  allow  amortization  of  wasting  assets,  but  does  allow  the  carrying  forward 
of  losses. 


1919-1929  829 

The  wage  bill1  includes  domestic  servants  but  excludes  all,  even  the 
lowest,  ranks  of  salaried  labor;  hence,  it  means  something  altogether 
different  from  both  the  American  and  the  German  series  used.  It  sub- 
stantially retained  its  over-time  relation  to  national  income,  displaying, 
however,  a  somewhat  more  pronounced  tendency  to  rise,  while  the  share 
of  profits  was  smaller  than  in  1911,  all  of  which  is  as  it  should  be  according 
to  Kondratieff  expectation.  Total  wages  (in  millions  of  pounds)  were 
1,399  in  1924;  1,437  in  1925;  1,382  in  1926;  1,492  in  1927;  1,479  in  1928; 
and  1,486  in  1929. 

Undivided  (total)  profits,2  which  with  admirable  freedom  from  preju- 
dice Mr.  Clark  recognizes  as  "the  principal  source  of  savings  under 
modern  conditions,"  displayed  a  falling  "trend."  The  highest  figure 
(186)  occurs  in  1924;  then  there  was  a  decline  in  1925  (169)  and  1926 
(134),  but  imperfectly  reversed  in  1927  (158)  and  1928  (same);  1929  (138) 
ushers  in  the  decline,  which  in  England,  however,  never  went  into  nega- 
tive figures  (minimum,  1931,  was  still  28).  But  again  we  find  that 
consumers'  outlay  comes  very  close  to  the  income  total.  Mr.  A.  E. 
Feavearyear3  has  made  the  attempt  to  estimate  average  annual  national 
expenditure  for  1924  to  1927.  If  we  add  to  his  list  of  items  of  consumers' 
expenditure  the  new  houses  and  the  furniture  which  he  includes  in  saving, 
we  get  about  3.7  billion  pounds,  which  is  above  any  of  our  income  figures 
for  those  years  except  for  1929.  In  spite  of  all  differences  in  conceptual 
arrangements,  Mr.  Clark  arrives  at  a  substantially  similar  conclusion, 
though  only  for  1929  (op.  tit.,  Table  112,  p.  252,  and  Diagram  IV  and 
comments).  In  fact,  "it  seems  clear  that  the  consumption  by  the  rich 
[and  "rich"  is,  according  to  Mr.  Clark,  everyone  with  an  income  exceed- 
ing 250  pounds]  was  in  1929  about  level  with  their  private  incomes." 
The  unavoidable  inference  is  that  there  cannot  have  been  much,  if  any, 
net  saving  by  private  households.  Repayments  of  building  loans  are 
nothing  else  but  payments  for  a  consumers'  good  bought  on  the  install- 
ment plan,  the  funds  of  the  building  societies  largely  represent  not  new 

1  Op.  tit.,  p.  28. 

2  Op.  tit.,  p.  187. 

3  Spending  the  National  Income,  Economic  Journal  for  March  1931,  see  table  on  p.  60. 
The  way  in  which  Mr.  Feavearyear  arrives  at  his  figure  for  savings,  which  is  not  less  than 
400  million  pounds,  is  a  good  example  of  the  way  in  which  the  usual  estimates  are  arrived 
at:  buying  a  piano  indicates  saving;  and  so  does  buying  a  house;  average  annual  amount 
of  new  capital  issues  accounts  for  256  millions — no  deduction  for  duplication  or  for  pay- 
ments from  ad  hoc  created  funds,  etc. — and,  of  course,  there  "must"  have  been  expansion  of 
private  business  out  of  income,  and  there  it  is.     Mr.  Feavearyear's  result,  presented  in 
another  paper  (Capital  Accumulation  and  Unemployment,  June  1936) — viz.,  that  the 
aggregate  value  of  private  fortunes  in  Great  Britain,  deflated  for  changes  in  prices  of 
assets,  has  up  to  1929  been  increasing  by  about  300  millions  a  year — though  arrived  at 
after  careful  investigation,  also  ceases  on  analysis  to  mean  what  it  seems  to. 


830  BUSINESS  CYCLES 

savings  but  shifts  in  investment.  "Security  savings"  by  workmen  and 
the  lower  middle  classes  and  net  accumulations  in  life  and  other  insurance 
companies  there  were,  of  course.  But  even  they — if  for  the  purpose  in 
hand  we  agree  to  treat  them  as  if  they  were  savings — must  have  been 
balanced  in  part  by  excess  consumption,  public  and  private.  We  arrive, 
therefore,  concerning  the  relation  between  national  income  and  con- 
sumers' outlay,  at  a  result  similar  to  those  in  the  American  and  German 
cases.  One  significant  difference  remains,  however.  Owing  partly  to 
the  absence  of  dazzling  hopes  for  the  future  and  partly  to*  what  the 
writer  has  no  other  words  for  than  character  or  moral  stamina,  there 
seems  to  have  been  in  England  no  such  general  rush  into  debt  as  there 
was  in  the  two  other  countries.1  And  the  credit  manufacturing  appara- 
tus proffered  much  less  temptation  to  it. 

e.  Profits  and  wages  call  for  additional  comment.  Concerning  the 
former,  see  for  United  States  corporate  profit  ratios — percentage  ratio 
of  net  income  minus  income  tax  to  gross  income — Chart  XLV;  for  United 
States  corporate  earnings  Chart  LIV;  for  German  dividends — as  rather 
doubtful  indicators  of  earnings — Chart  XL VIII;  and  for  United  Kingdom 
"profits"— the  most  doubtful  figures  of  all— Chart  XLIX.  We  will 
confine  our  comments  to  this  country,  because  American  research  on  the 
subject,  much  superior  to  any  other,  has  most  nearly  succeeded  in  bring- 
ing out  the  main  contours  of  this  complicated  tangle  of  facts.2 

In  spite  of  the  excellent  work  done  we  are,  however,  as  yet  far  from 
seeing  clearly,  and  any  inference  has  to  contend  with  disheartening 
margins  of  error.  Not  only  is  the  raw  material  the  product  of  book- 
keeping processes  which  unavoidably  deviate,  to  an  unknown  extent 
that  is  sure  to  vary  as  between  industries  and  individual  concerns,  from 

1  Housing  might  be  considered  an  exception;  but  subsidized  and  safeguarded  and  run 
on  the  your-rent-will-buy-your-house  plan  as  it  was,  it  may  fairly  be  looked  upon  as  a 
special  case. 

2  The  author's  primary  obligation  is  to  the  work  of  Professor  Crum,  in  particular,  to 
Corporate  Earning  Power,  1929,  but  also  to  his  many  papers  on  the  subject.     In  the  second 
place,  Professor  R.  C.  Epstein's  great  investigation  has  been  of  great  help:  Industrial 
Profits  in  the  United  States,  National  Bureau  Publication,  1984,  with  an  introduction  by 
Professor  Mitchell.     But  on  the  theory  brought  to  bear  on  the  interpretation  of  the  data  cf. 
Mrs.  Tappan  Hollond's  review  in  the  Economic  Journal,  1935.     In  the  third  place,  mention 
should  be  made  of  L.  H.  Sloan  Corporation,  Profits,  1929;  S.  H.  Nerlove,  A  Decade  of 
Corporate  Incomes,  1932;  R.  T.  Bowman,  The  Statistical  Study  of  Profits,  1934,  W.  A. 
Paton,  Corporate  Profits  as  Shown  by  Audit  Reports,  National  Bureau  Publication,  1935; 
and  Professor  F.  C.  Mill's  Economic  Tendencies;  as  well  as  to  a  number  of  papers  by  L. 
Bagwell,  L.  R.  Robinson,  and  others.     The  only  way  toward  real  insight  would,  of  course, 
be  the  detailed  study  of  the  life  history  of  individual  concerns,  part  of  the  material  for 
which  is  to  be  found  in  the  discussions  of  annual  statements  in  the  financial  press.     For 
Germany  the  Bilanzanalysen  of  the  Deutsche  Volkswirt  are  an  excellent  source  of  this  type 
of  information. 


1919-1929  831 

the  actual  state  of  things;  and  not  only  does  it  largely  fail  to  represent  at 
all  adequately  the  limbo  in  which  dwell  the  abortive  and  short-lived 
attempts  at  enterprise;  but  even  if  that  were  not  so,  profit  figures  would 
at  their  best  give  but  a  medley  of  economically  heterogeneous  elements, 
of  which  entrepreneurial  profits1  are  only  one,  though  we  may  perhaps 
hope  that,  being  the  most  active  one,  they  will  show  up  better  in  the 
fluctuations  of  the  aggregate  than  on  other  grounds  we  have  a  right  to 
expect. 

Since  these  difficulties  particularly  interfere  with  the  meaning  of  the 
various  ratios  that  have  been  computed,  we  had  better  start  with  aggre- 
gate net  income  of  all  corporations  before  payment  of  income  taxes  and 
dividends.2  The  violent  fluctuations  at  the  beginning  are  due  to  war 
effects  and  to  the  Juglar  depression  (1919,  9.3  billions;  1921,  0.64),  but 
from  1923  (6.64)  to  1929  (9.13)  we  have  what  looks  like  an  almost 
steadily  rising  "trend,"  substantially  similar  to  that  in  total  national 
income:  the  per  cent  relation  of  corporate  income  to  realized  national 
income  (Copeland-Crum  series)  plus  corrected  corporate  accumulations, 
was  fairly  constant — for  1923  to  1929  it  was  9.2,  8,  10.2,  9.5,  8.3,  10, 
10.3.  Within  this  contour,  cyclical  phases  are  clearly  recognizable  both 
in  absolute  figures  and  in  percentages.  The  rise  of  the  fourth  Juglar  is 
particularly  well  marked:  from  1924  (5.74  billions)  there  is  a  character- 
istic jump  to  1925  (nearly  8  billions)  and  after  that  some  tapering  off 
(1926,  7.84  billions;  1927,  6.84),  interrupted  by  the  abnormalities  of 
1928  (8.67)  and  1929.  But  this  shows  what  we  are  to  think  of  that 
"trend."  On  the  one  hand,  1923  and  1924  belong  to  a  Juglar  recovery 
that  is  followed  by  a  Juglar  prosperity;  on  the  other  hand,  the  two 
abnormally  prosperous  years  happen  to  be  the  last  ones  of  the  series.3 

Of  the  other  ratios — we  do  not  revert  to  the  profit  ratio  in  Professor 
Crum's  sense — the  lowest  is  the  earnings  ratio  or  percentage  of  net  income 

1  As  we  have  seen  in  Chap.  Ill,  this  element  tends,  in  every  individual  case,  to  converge 
toward  zero,  but  does  not  otherwise  harbor  any  tendency  toward  equalization. 

2  Exclusive  of  tax-exempt  and  life  insurance  companies.     See  Ebersole,  Burr,  and  Peter- 
son in  Review  of  Economic  Statistics  for  November  1929,  and  Fabricant,  Recent  Corporate 
Profits  in  the  United  States,  National  Bureau  of  Economic  Research,  for  Apr.  18,  1934. 
Dividends  received  from  other  corporations  are  excluded. 

3  It  is,  hence,  inadmissible  to  use  that  "trend"  as  an  indication  of  long-run  tendencies 
inherent  in  capitalist  evolution  and,  in  particular,  as  an  indication  of  long-run  tendencies 
in  distributive  shares.     Even  disregarding  the  necessity  of  correcting  totals  for  increase  in 
investment,  the  true  picture  emerges  only  if  the  subsequent  depression  be  included.     But 
it  is  wholly  misleading  to  use  the  "trend"  in  total  dividend  payments — let  alone  dividends 
plus  cash  value  of  rights— for  either  purpose,  because,  as  stated  before,  they  absorbed  an 
increasing  percentage  of  total  corporate  net  incomes.     This  is  interesting  evidence  about 
the  alleged  saving  propensities  of  the  period,  but  is  otherwise  a  purely  intracapitalistic 
affair  which  has  no  bearing  on  the  relative  fortunes  of  social  classes  in  the  distributive 
process. 


BUSINESS  CYCLES 

to  total  assets,  which,  since  not  all  corporations  file  balance  sheets,  is  a 
matter  that  involves  much  estimating.  For  the  three  years  for  which 
Professor  Crum  calculated  it,  1924  to  1926,  it  was,  taking  all  divisions 
together,  roughly  between  2  and  3  per  cent,  manufacturing  industry 
leading  with  about  5.  For  1926  in  particular,  the  aggregate  return  on 
total  assets  was  1.98  per  cent,  for  manufacturing  industry  alone,  4.36.1 
The  latter  figure  just  about  equals  the  yield  of  highest  grade  bonds  in 
the  same  year — a  fact  which  is  of  some  importance  although  what  we 
have  before  us  is  an  average  and  not  a  marginal  quantity  and  although 
profit  and  interest  are  obviously  not  independent  of  each  other.  Now, 
whatever  ratio  may  be  relevant  for  other  purposes,  this  one  is  relevant  for 
ours.  And  since  it  is  quite  as  low  as  we  should  expect  it  to  be  in  the 
downgrade  of  a  Kondratieff,  we  conclude  that  there  cannot  have  been 
much  of  a  general  "profit  inflation,"2  for  this  would  have  shown  precisely 
in  the  general  level  of  income  per  unit  of  assets.  This  may  have  been 
different  in  1928  and  1929,  but  such  indications  as  we  have  do  not  sug- 
gest that  the  earnings  ratio  was  appreciably  higher.  In  some  cases  it  was 
lower  than  in  1926.  The  reason  why  it  was  so  low  is  of  course  that  all 
losses  made  by  reporting — i.e.,  still  existing — corporations  enter  into  it, 
as  well  as  the  gains.  Taking  only  corporations  reporting  gains,  Professor 
Crum  obtains,  for  1926,  6.95  per  cent  in  the  manufacturing  division  and 
3.66  in  all.  But  according  to  his  showing,  almost  half  of  all  corporations 
worked  at  a  loss  or  at  no  profit  or  at  practically  none.3  It  follows  that 
expectation  from  our  model  is  verified,  not  only  as  to  the  size  of  the 
earnings  ratio,  but  also  as  to  the  reason  for  that  size.  For  it  is  obvious 
that  such  prevalence  of  losses  or,  in  a  business  sense,  inadequate  returns, 

1  Professor  Crum  does  not  add  interest  charges:  as  stated,  he  considers  net  income. 
We  have  done  the  same  in  the  preceding  paragraph.     In  fact,  it  is  open  to  question  whether 
to  add  interest  (or  only  interest  on  long-term  debt)  is  conducive  to  a  correct  impression 
of  the  rate  of  profits.     It  might  just  as  well  be  argued  that  what  we  ought  to  do  is  to  deduct 
also  interest  on  owned  capital. 

2  That  term  is  not  used  in  Mr.  Keynes'  sense,  Treatise  on  Money  I,  p.  155,  but  merely  to 
indicate  all  cases  in  which  profits,  by  virtue  of  being  obviously  abnormal,  might  be  taken 
as  a  symptom,  consequence,  or  cause  of  other  abnormalities  or  disturbances,  e.g.,  as  a 
cause  of  overexpansion  of  output  or  investment  or  as  a  symptom  of  monopoloid  under- 
utilization  of  resources. 

3  This  finding  is  substantially  confirmed  for  the  whole  period  by  Professor  Epstein's 
investigation,  see  op.  cit.,  p.  457:  among  manufacturing  corporations  the  "with  net  income" 
groups  amounts  to  roughly  60  per  cent  of  the  total  which  was,  1919  to  1928,  between 
70,000  and  90,000.     "Thus  in  all  years  from  1919  to  1928  the  number  of  manufacturing 
corporations  with  net  incomes  runs  from  about  50,000  to  55,000  a  year,  except  in  1921  when 
the  figure  was  slightly  less  than  40,000."     Hence,  if  Professor  Epstein  rightly  takes  excep- 
tion to  the  "common  impression  •  •  •  that  about  50  per  cent  of  the  corporations  in  the 
country  lose  money,"  we  ought  to  add  that  this  common  impression  expresses  a  very 
important  truth  much  more  nearly  correctly  than  it  is  usual  for  common  impressions  to  do. 


1919-1929  833 

is  but  another  symptom1  of  the  vigor  with  which  our  competing-down 
process  did  its  work  and  of  the  fact  that  the  prewar  process  still  persisted : 
from  our  standpoint,  all  this  was  entirely  normal  and  merely  one  of  the 
aspects  of  economic  "progress."2  If  it  were  possible  to  go  into  details, 
further  conformity  with  expectation  would  reveal  itself.3 

Next  we  will  glance  at  rates  of  return  on  book  value  of  stock  equity 
of  "all"  corporations.  Mr.  Fabricant's  figures4  for  1927  to  1930  will 
suffice.  In  the  grand  total  they  are  5.3,  6.2,  6.2,  2.2;  for  manufacturing 
alone  6.2,  7.6,  8.3,  2.6  per  cent.  They  differ  from  earnings  ratios  by 
virtue  of  the  agreements  entered  into  with  one  another  by  the  various 
groups  of  capitalist  claimants,  and  they  have,  hence,  for  our  purpose, 
little  importance  in  themselves.  But  they  may  again  be  compared  to 
bond  yields.  Discarding  rather  revealing  details,  we  again  get  the 
impression  that  under  this  aspect  even  these  rates  do  not  suggest  "profit 

1  There  were  of  course  corporations  which  existed  for  the  very  purpose  of  working  at  a 
loss;  others,  the  profits  of  which,  or  more  than  that,  went  to  the  executives;  and  still  others 
which,   being  subsidiaries  of  some  other  concern,  were  on  principle  selling  at  cost  or 
even  at  a  loss.     But  it  will  hardly  be  averred  that  cases  of  these  types  were  significant 
enough  to  interfere  with  our  conclusions.     Nor  can  it  be  objected  that  the  evolutionary 
process  failed  to  eliminate  the  antiquated,  ill-conceived,  or  otherwise  unsuccessful  elements. 
We  have  in  earlier  chapters  dwelt  on  the  reasons  why  losing  concerns  often  "hang  on" 
for  some  time.     But  the  going-out-of-business  rate  was  considerable  throughout,  only 
it  does  not  show  in  this  kind  of  material.     Finally,  it  cannot  be  objected  that  the  process 
failed  to  work  in  the  sphere  of  bigger  and  of  big  corporations  which  were  individually  suc- 
cessful more  or  less  all  the  time.     For  the  relation  between  size  and  success,  although 
not  one  of  proportionality,  works  not  only  one  way.     It  may  be  added  that,  since  the 
competing-down  process  takes  time  (if  it  did  not,  economic  life  would  be  a  continuum  of 
catastrophes)  we  shall  not  share  Professor  Epstein's  astonishment  at  finding  that  sub- 
stantially the  same  divisions  and  concerns  were,  throughout,  at  or  near  the  watershed  of 
loss:  all  that  this  proves  is  that  business  life  is  not  a  game  of  chance  and  that  profits  are 
not  adequately  described  as  windfalls. 

2  What  was,  from  the  standpoint  of  our  analysis,  a  highly  "normal"  state  of  things  does 
not  seem  so  to  other  economists.     Professor  Crum  seems  to  consider  the  facts  that  "num- 
bers of  enterprises  •  •  •  are  dragging  along  with  a  very  low  rate  of  return  on  their  prop- 
erty," and  that  "a  considerable  share  of  the  total  gross  corporate  business  is  done  at  a  loss" 
as  a  reason  to  doubt  "the  long-run  healthfulness  of  corporate  industry."     It  is  submitted 
that  attention  to  the  logic  and  rhythms  of  the  capitalist  process  completely  removes  any 
such  doubt,  although  that  eminent  economist  was  perfectly  right  if  he  intended  his  guarded 
statement  to  imply  a  prediction  of  impending  vicissitudes.     Other  economists  are  in  the 
habit  of  taking  the  facts  discussed  as  proof  of  the  wastefulness  of  competition  or  of  capital- 
ism in  general.     If,  however,  they  mean  more  than  a  triviality,  they  are  wrong.     Those 
losses  are  not  waste  or  per  se  indications  of  waste  in  the  sense  that  the  social  organism  as 
a  whole  gets  nothing  in  return,  provided  they  are  placed  in  their  proper  setting.     And 
they  would,  of  course,  be  unavoidable  also*  in  a  planned  economy. 

3  Especially  in  a  study  of  the  concerns  that  enjoyed  more  than  average  returns.     Mr. 
Sloan's  study   (op.  dt.)  of  545  big  corporations  sheds  some  light  on  this  and  is  recom- 
mended to  the  reader. 

4  Op.  cit.,  p.  8;  net  income  is  taken  "after  tax." 


834  BUSINESS  CYCLES 

inflation."  In  1927  AAA  bonds  yielded  a  little  over  4.3  per  cent,  in 
1929  nearly  5.  European  prewar  experience — American  bookkeeping 
methods  before  1909  do  not  permit  comparison  with  American  prewar 
experience — would,  very  roughly,  support  the  view  that  such  a  margin 
was  about  normal.  And  the  1930  figure  is  also  "normal"  for  incipient 
depression.  Textiles,  leather,  and  rubber  display  decline  from  1927  to 
1929;  food-beverages-tobacco,  paper  and  pulp,  stone-clay-glass  are  exam- 
ples of  comparative  stability;  chemicals  and  metals,  of  strong  increase 
• — surely  nothing  to  be  surprised  at. 

Finally,  Professor  Epstein's  figures  for  percentage  profit  (net  income 
plus  interest  on  funded  debt,  "before  tax")  to  total  capital  (stock,  com- 
mon and  preferred,  surplus,  undivided  profits,  funded  debt,  but  not  other 
debt)  are  for  1924  to  1928 11  5.9,  7.5,  7.2,  6.4,  7.3,  but  for  the  sample  of 
3,144  corporations,  9.2,  10.7,  10.9,  9.4,  10.2.2  The  latter  may  serve  to 
illustrate  what  we  should  have  taken  as  evidence  of  abnormally  high 
profits  if  this  had  been  the  result  for  all  corporations  or  even  a  random 
sample.  The  2,046  manufacturing  corporations  display  still  higher 
figures  but,  for  1926,  the  modal  percentages  are  from  5  to  9  and  the  high 
average  is  as  much  due  to  the  inclusion  of  all  peak  successes  as  to  the 
almost  complete  absence  (less  than  4  per  cent)  of  cases  of  loss.  Cyclical 
fluctuations,  the  rise  of  the  fourth  Juglar  in  particular,  are  much  in 
evidence :  it  is  interesting  to  notice  that,  contrary  to  a  prevalent  impres- 
sion, 1925  (or  1926)  was  the  most  profitable  year,  though,  of  course, 
overshadowed  by  1919  and,  before  that,  by  1916-1917.  The  range  of 
variation  as  between  the  106  constituent  groups  and,  within  them,  as 
between  concerns  is  as  wide  as  we  should  expect.3 

/.  Gathering  up  the  threads  of  our  analysis  of  wages  and  employ- 
ment (this  section,  I,  a,  II,  a,  and  II,  d;  Sec.  C  and  Sec.  E9  passim; 
Chap.  XI)  we  will  now  discuss  the  behavior  of  wage  rates4  and  the  effects 

1  Op.  cit.  p.  50. 

2  Op.  cit.,  p.  53. 

8  The  limited  purpose  of  this  sketch  makes  it  impossible  to  do  justice  to  the  rich  lode 
of  information  to  which  Professor  Epstein's  work  has  opened  access.  It  must,  however, 
be  stressed  again  that  there  is  no  reason  why,  in  a  period  such  as  1919  to  1929,  and  in  a 
sample  constituted  as  this  is,  extra  gains  should  balance  extra  losses.  Nor  is  it  easy  to 
see  why  Professor  Epstein  should  so  categorically  aver  that  the  differences  in  earning  rates 
of  different  industries  "cannot  be  regarded  as  differentials  due  to  the  rent  of  superior  busi- 
ness abilities"  (p.  582).  Nothing  is  further  from  the  present  writer's  mind  than  a  wish 
to  defend  that  antediluvian  turn  of  phrase,  which  moreover  points  in  the  wrong  direction. 
But  in  looking  at  Professor  Epstein's  list  of  particularly  successful  industries,  one  may  well 
wonder  whether  success  has  not  very  much  to  do  with  the  quality  of  the  products  of  some 
and  the  dash  of  the  advertising  of  others.  And  what  is  there  so  very  "unlikely"  in  the 
statement  that,  on  the  one  hand,  able  men  take  to  promising  jobs  and,  on  the  other  hand, 
jobs  become  promising  in  the  hands  of  able  men? 

*  We  use  rates  wherever  possible  because  they  come  nearest  to  representing  the  price 


1919-1929 


835 


that  may  have  emanated  from  them.     American  facts  are  presented  in 
Chart  XL VII. 

1.  In  appraising  the  evidence  contained  in  these  curves  a  number  of 
limitations  must  be  borne  in  mind  which  impose  extreme  caution  in 


1919     1920    1921      1922    1923    1924  _  1925  .  1926 .  1927    1928  _  1929    1930    193L_  1932  _  1933    1934    1935 
CHART  XL VII. — United  States  (see  Appendix,  p.  1070). 

drawing  inferences.  Pay  rolls,  which  have  been  plotted  again  and  can 
now  be  compared  with  both  " corrected J>1  and  "real"  pay  rolls,  and 
employment  represent,  of  course,  pay  rolls  and  employment  in  manufac- 
turing industry  only.2  Hence,  even  disregarding  that  they  do  so  imper- 

of  a  definite  quantity  of  labor.  Weekly  earnings  per  workman  or  per  employed  workman 
or  per  working-class  family  are,  of  course,  more  important  for  considerations  about  welfare, 
social  justice,  and  so  on.  They  would  be  more  important  than  rates  also  for  us  if  we  could 
enter  into  the  question  how  the  working  class  fared  during  what  was  in  this  country  a 
span  (though  not  one  which  can  be  taken  as  either  typical  or  near  average)  of  relatively 
uninhibited  capitalism.  This  question  must,  however,  remain  outside  of  our  range. 
We  will  merely  note  that  in  this  country  actual  weekly  earnings  in  manufactures,  which 
had  risen  to  a  peak  of  .$29.48  in  1920,  suffered  a  "normalizing"  decline  to  $23.23  in  1922 
and  then  rose  again  to  within  10  per  cent  of  that  peak  ($27.36)  in  1929.  See  L.  Wolman, 
National  Bureau  Bulletin  for  May  1,  1933,  p.  2. 

1  The  same  terminology  is  used  in  this  argument  as  in  our  discussion  of  prewar  wages  and 
employment. 

2  For  a  closer  analysis,  see  W.  A.  Berridge,  Review  of  Economic  Statistics  for  November 
1930. 


836  BUSINESS  CYCLES 

fectly  (as  is  obvious  from  the  description  in  the  appendix),  they  do  not 
indicate  the  course  of  total  income  and  total  employment  of  the  working 
class  but  only  what  happened  in  a  sector  of  the  national  economy.  The 
full  importance  of  this  becomes  evident  if  we  recall  that  the  develop- 
ments in  manufacturing  industry  were  instrumental  in  creating  additional 
employment  and  labor  income  in  certain  other  sectors — the  sector  of 
"services"  in  particular — so  that,  in  comparing  manufacturing  pay  rolls 
with  say  value  of  output  of  manufactures  or,  as  we  did  before,  with 
total  national  income,  we  are  not  isolating  a  self-contained  relation  but 
cutting  through  a  nexus  which  is  at  the  same  time  more  comprehensive 
and  more  relevant.1  Moreover,  all  the  series  used  are  open  to  objection 
on  various  counts,  that  of  cost  of  living,  in  particular,  to  the  old  one  that 
the  improvement  in  quality  of  the  commodities  which  entered  the  Ameri- 

1  Uncritical  comparisons  of  wage  bills  or  wage  rates  with  other  aggregates,  often  used  as 
a  basis  for  inferences  and  value  judgments  about  distributive  shares,  are  a  frequent  source 
of  much  coarser  errors  than  the  one  alluded  to  in  the  text,  if  not  completely  meaningless. 
For  instance,  there  is,  for  the  purposes  of  the  study  of  cyclical  fluctuations,  some  sense  in 
comparing  fluctuations  in  wage  bill  with  fluctuations  in  the  Census  Bureau  figures  of  total 
transactions  in  manufactured  products  which  is  sometimes  misleadingly  referred  to  as  the 
gross  value  of  those  products  and  differs  from  value  added  plus  value  of  domestic  raw 
materials  plus  value  of  imported  producers'  goods  mainly  by  the  interfirm  transactions  in 
commodities.  But  there  is  no  sense  in  instituting  this  comparison  for  the  purpose  of 
measuring  labor's  share.  Nor  is  there  any  sense  in  comparing  average  per  capita  earnings 
of  employees  with  the  sum  total  of,  say,  stockholders'  cash  receipts  or,  still  worse,  with 
stockholders'  cash  receipts  plus  stock  dividends,  or  with  dividends  plus  undivided  surplus. 
Much  erroneous  argument  is  due  to  the  confusion,  adverted  to  in  Chap.  II,  of  product  per 
man-hour  and  the  productivity  of  labor.  This  accounts  for  the  habit  of  speaking  of  varia- 
tions in  the  former  as  if  they  were  variations  in  the  "social  contribution"  of  labor,  and  of 
comparing  them  with  variations  in  wage  rates  on  the  hypothesis  that  proportional  covaria- 
tion between  the  two  is  in  some  sense  normal  and  that  deviations  from  it  are  something  to 
be  wondered  at  or  to  criticize.  We  are  not  attacking  any  ideals  that  may  be  the  premises, 
but  the  economic  errors  that  are  at  the  basis,  of  judgments  about,  and  sometimes  even  of 
mere  presentations  of,  the  facts — errors  which  very  simple  considerations  on  Boehm-Bawer- 
kian  lines  should  be  sufficient  to  dispel.  The  boldest  attempt  at  introducing  econqmic 
meaning  into  discussions  of  statistics  of  that  kind  has  been  made  by  Professor  P.  Douglas. 
None  of  the  many  reservations  that  have  to  be  made  on  many  counts,  detracts  from  the 
merit  of  his  Theory  of  Wages,  1984.  It  should  be  added,  however,  that  many  economists 
who  do  not,  in  dealing  with  wage  questions,  forget  their  courses  in  elementary  theory, 
yet  fail  to  realize  the  importance  of  the  fact  that  the  fundamental  theorem  about  marginal 
value  productivity  of  labor  is  an  equilibrium  proposition  that  would  at  best  apply  (approxi- 
mately) in  neighborhoods  of  equilibrium,  but  cannot  in  the  intervals  between  them. 
Profits  in  our  sense  precisely  arise  and  vanish  in  these  intervals,  hence,  do  not  bear  any 
definite  relation  to  "productivity  wages."  Moreover,  if  the  reader  will  forgive  a  triviality 
which  is  always  obscured  in  popular  discussions,  variations  in  productivity  wages  and  the 
corresponding  variations  in  output  have  much  less  to  do  with  what  in  common  parlance  we 
understand  by  the  personal  efficiency  of  the  workmen  or  with  what  in  this  sense  may  be 
attributable  to  their  efforts,  than  with  the  variations  in  the  amounts  of  the  other  factors 
applied  and  with  the  changes  in  "methods  of  production." 


1919-1929  887 

can  working-class  budget — food,  clothing  and  what  we  have  termed 
gadgets  of  modern  life  in  particular — was  one  of  the  outstanding  features 
of  the  period.  Finally,  wage  rates  varied  so  greatly  between  localities 
and,  in  the  same  place,  between  industries  and  firms  that,  as  we  have 
seen  in  an  earlier  chapter,  speaking  of  one  national  wage  rate  and  its 
variations  might  well  be  thought  inadmissible,  even  if  the  concept  of  a 
national  rate  per  hour  were  unexceptionable  in  itself.  But  it  is  not. 
Piece  rates,  bonuses,  and  so  on  drive  a  wedge  between  hourly  rates  and 
hourly  earnings,  which  makes  it  impossible  to  infer  the  one  from  the 
other,  still  more  rates  per  hour  from  earnings  per  week  or  pay  rolls. 
In  using  the  terms  hourly  earnings  and  hourly  rates  as  if  they  were  syn- 
onymous, we  are  guilty  of  a  serious,  though  very  common,  misdemeanor, 
in  extenuation  of  which  we  can  only  plead  that  this  difficulty  was  not  so 
great  from  1923  to  1929  as  it  afterwards  became. 

Expectation  as  to  wage  rates  in  the  Kondratieff  phases  obtaining 
during  the  period  is  for  a  moderate  rise  in  money  and  a  substantial  one 
in  corrected  and  real  rates.  If  we  start  'from  the  middle  of  1923,  this 
is,  on  the  whole,  what  we  find,  and  the  Juglar  and  Kitchin  variations 
are  also  recognizable.1  In  other  words,  there  is,  as  far  as  that  piece 
of  evidence  goes,  no  reason  to  suppose  that  variations  in  wage  rates 
were  anything  but  normal  in  the  sense  that  they  did  not  interfere  with 
the  expansion,  or  contribute  to  the  subsequent  contraction,  of  business 
volumes  by  being  "too  high":  the  only  question  seems  to  be  whether 
they  were  not  "too  low."  Again,  however,  we  observe,  surveying  the 
whole  period  and  comparing  the  level  at  which  rates  moved  with  that 
of  the  last  prewar  years,  not  only  the  traces  of  the  monetary  revolution, 
as  we  do  in  other  comparable  quantities,2  but  also  the  fact  that  the  down- 
ward revision  after  the  postwar  peak  was  comparatively  small  and  only 
temporary.  Hence,  it  is  at  least  a  possibility  that  the  war  and  the 
postwar  boom  left  labor  a  relatively  dearer  factor  of  production  than  it 
had  been  before.  That  it  became  a  relatively  more  expensive  consumers' 
good  is,  of  course,  beyond  reasonable  doubt.  Suggestion  to  that  effect 
is  present  in  nearly  all  available  wage-rate  series,  but  we  will  content 

1  Better  than  in  the  series  used  in  our  chart  (see  Appendix)  the  course  of  wage  rates 
shows  in  the  new  composite  wage  index  of  the  Federal  Reserve  Bank  of  New  York,  the 
latest  improvement  of  Mr.  Snyder's  index  (see  Monthly  Review  of  Credit  and  Business 
Conditions,  for  Feb.  1,  1938,  p.  12,  where  that  comprehensive  index,  1926  =  100,  has  been 
charted  on  a  natural  scale).     The  increase  in  money  wage  rates  between  1923  and  1929 
is  quite  appreciable,  the  rise  of  the  fourth  Juglar  shows  well.     This  should  be  noticed, 
because  there  is  a  prevalent  impression  to  the  effect  that  from  1923  rates  did  not  rise  at  all. 

2  Price  levels,  contrary  to  a  widely  held  opinion,  are  not  among  those  comparable 
quantities.     The  reader  recalls  that  expectations  for  commodity  prices  and  wages  differ 
fundamentally  and  that  there  is  not  the  same  reason  to  expect  that  wages  should  fall  in 
recession  as  there  is  that  prices  should. 


838  BUSINESS  CYCLES 

ourselves  with  one,  the  basic  wage  rate  for  common  labor  in  the  Pitts- 
burgh district,  as  reported  by  the  U.  S.  Steel  Corporation.1  This  rate 
was  slightly  under  20  cents  per  hour  for  1913  (monthly  average)  and 
nearly  51  for  1920.  It  declined  in  1921  and  part  of  1922,  but  was  back 
again  to  50  cents  in  the  monthly  average  of  1924,  after  which  it  remained 
constant  through  1930. 2  It  does  not  make  a  great  deal  of  difference 
whether  we  say  that  the  irregularities  at  the  threshold  of  our  period  veil 
the  extent  of  the  rise  in  money  wage  rates  or,  as  we  put  it  above,  that 
they  raised  the  level  from  which  the  "regular"  developments  start. 
In  both  cases,  the  course  of  wages  must  be  considered  as  resulting  from 
two  component  tendencies — one  which  tended  to  depress  and  one  which 
tended  to  raise  them.  On  the  whole,  however,  the  latter  way  of  express- 
ing American  wage  facts  seems  preferable  for  our  purpose,  because  it 
brings  out  more  clearly  that  that  "level,"  created  by  war  and  postwar 
irregularities  and,  owing  to  the  resistance  offered  by  the  environment  to 
downward  revision,  substantially  left  as  it  was,  was  a  noncyclical  and 
almost  constant  force  throughout. 

2.  Even  if  we  look  at  wage  rates  in  this  light,  we  shall  hardly  find 
any  reason  to  recede  from  the  result  provisionally  stated  above,  viz., 
that  wages  did  not  hamper  prosperities  or  cause  relapses  by  being  "too 
high."  There  is  neither  any  evidence  the  writer  can  think  of  that  they 
did  nor  any  expectation  to  that  effect  from  theory — within  limits,  a 
wage  level  that  persists  through  a  decade  becomes  a  datum  to  which  the 
system  will  in  general  adapt  itself  without  changing  its  mode  of  working.3 

1  The  series  has  much  to  recommend  it  but  is  not  quite  consistent  in  meaning.     In 
particular,  these  rates  applied  to  working  days  of  different  length,  which  is  not  only  relevant 
to  daily  earnings,  and  1918  to  1921  also  reflect  the  higher  payment  for  overtime.     The 
series  of  hourly  rates  paid  for  common  labor  in  road  building  (Bureau  of  Public  Roads) 
gives  a  somewhat  different  picture.     In  1915  the  rate  was  20  cents.     It  steadily  rose  to 
49  for  1920,  declined  in  1921  and  1922  to  82.     Then  it  was  at  38  cents  for  3  years  and  rose 
in  1928  (maximum)  only  to  40.     Hourly  wages  in  soft-coal  mining  continued  to  fall  after 
1922  (85.3  cents;  1929,  65.9). 

2  The  index  referred  to  in  the  last  note  but  two  rises  spectacularly  in  1919  and  (three 
quarters  of)  1920,  then  falls  sharply  through  1921  and,  though  recovering  substantially 
in  1922  and  especially  1923,  does  not  come  as  near  to  the  all-time  peak  as  the  rate  discussed 
in  the  text.     On  the  other  hand,  it  continues  to  rise  and  ends  up  in  1929  with  a  value  higher 
than  that  peak.     The  implication  is,  hence,  fundamentally  the  same. 

3  Qualifications  of  this  proposition  will  be  evident  from  the  rest  of  the  paragraph. 
It  would  not  be  true,  of  course,  of  all  deviations  from  the  "normal"  course  of  things.     But 
it  should  be  noticed  that  to  some  extent  it  agrees  with  an  opinion  that  is  at  present  held  by 
many  economists,  who  would,  however,  neither  stress  the  long-run-level  aspect  nor  accept 
the  qualifications.     Professor  Myrdal's  and  Mr.  Keynes'  teaching  may  be  referred  to  in 
illustration.     To  put  it  differently,  the  above  proposition  formulates  an  element  of  truth 
that  is  contained  in  the — otherwise  untenable — opinion  that  the  absolute  level  of  monetary 
wage  rates  does  not  matter.     There  is  another  element  of  truth  in  it  which  is,  however, 
trivial:  changes  in  the  absolute  level  of  monetary  wage  rates  do  not  matter  if  all  other 
monetary  magnitudes  and  expressions  move  uno  actu  and  proportionately. 


1919-1929  839 

But  this  merely  means  that  the  general  complexion  of  successive  business 
situations  was  not  substantially  affected  by  it,  i.e.,  not  only  that  pros- 
perity remained  prosperity  and  recession  remained  recession,  but  also 
that  all  phases  were  presumably  as  "intensive"  as  they  would  have  been 
with  a  somewhat  lower  level  of  wages.  It  does  not  mean  that  there  were 
no  effects,  in  particular  on  employment.  For  one  of  the  ways  in  which 
the  system  would  adapt  itself  to  a  high  level  of  wages  consists  precisely 
in  making  of  the  dear  labor  factor  or  labor  commodity  as  economical 
a  use  as  possible.  We  shall,  therefore,  suspect  that  during  the  period 
an  additional  source  of  unemployment  may  have  been  present  to  swell 
the  total:  we  have  already  noticed  that  the  Kondratieff  phases  into  which 
our  period  falls,  would  "naturally'*  display  considerable  technological 
unemployment;  moreover,  it  goes  without  saying  that,  especially  in 
1921,  but  also  in  other  years — 1927,  for  instance — there  must  have  been 
cyclical  disturbance  unemployment  of  the  kind  that  is  not  directly 
traceable  to  innovation,  as  well  as  unemployment  due  to  random  causes, 
such  as  the  Mississippi  flood  and  others;  we  may  now  have  to  add  unem- 
ployment of  the  type  which  we  have  called  vicarious. 

Now,  our  factual  knowledge  and  analytic  powers1  being  what  they 

1  We  are,  moreover,  at  a  disadvantage  as  against  such  theoretical  reasoning  on  the 
subject  as  has  recently  been  offered,  since  we  cannot  state  the  theoretical  case  without 
expanding  this  subsection  into  a  treatise  on  wages.  But  it  should  be  observed  that  the 
above  argument  about  economizing  dear  factors  is  not  open  to  the  objection  that  any 
increase  in  the  expenses  of  production  which  is  due  to  increase  in  wages  is  at  least  com- 
pensated by  an  equal  increase  in  producers'  revenue.  For,  even  if  that  were  always  so, 
the  individual  firm  would  still  have  a  motive  to  react  to  an  increase  in  wages  by  a  reduction 
of  the  labor  employed  per  unit  of  product.  The  consequent  rearrangement  of  its  combina- 
tion of  factors  cannot  be  neglected  for  a  period  of  the  length  of  a  Juglar,  or  even,  in  many 
cases,  of  a  Kitchin.  And  this  is  as  true  under  conditions  of  imperfect  competition  as  it  is 
under  conditions  of  perfect  competition.  It  should  be  added  that  in  other  respects  also 
prevalence  of  conditions  of  imperfect  competition  affects  the  argument  about  the  conse- 
quences on  employment  (and  output)  of  variations  in  monetary  and  real  wage  rates  less 
than  might  be  thought.  The  elegant  argument  presented  by  Dr.  P.  Sweezy  at  the  1937 
meeting  of  the  American  Economic  Association  (see  American  Economic  Review,  Supple- 
ment, March  1938,  p.  156)  may  serve  as  an  example.  The  present  writer  entirely  agrees 
with  his  emphasis  on  the  importance  for  employment  of  cyclical  shifts  in  the  demand  curve 
for  products  (p.  157)  and  has  been  at  pains  to  stress  this  (Chaps.  XI  and  XII)  both  in  the 
matter  of  wages  and  of  interest  rates.  It  is  partly  for  this  reason  that  the  short-run 
influence  on  cyclical  phases  of  variations  in  wage  rates  has  not  been  assigned  a  more 
important  role  in  the  text.  But  so  far  as  Dr.  Sweezy's  argument  is  based  on  the  proposition 
that  with  imperfect  competition  firms  will  not  react  to  shifts  in  their  cost  curves,  because 
this  involves  raising  the  prices  of  their  product,  which  would  drive  business  away,  or 
lowering  them,  which  is  self-defeating  since  it  will  induce  competitors  to  follow  suit,  it  has, 
while  logically  correct,  little  claim  to  being  considered  more  realistic  than  others.  For  it  is 
obviously  more  likely  that  competitors  of  the  firm  that  contracts  output  and  raises  prices 
because  of  an  increase  in  wages,  will  do  the  same  than  it  is  that  they  will  try  to  conquer  the 
field  from  which  the  latter  now  retires;  and  it  is  no  less  obvious  that  in  a  highly  "dynamic" 
society  the  motive  for  expanding  output  and,  for  this  purpose,  reducing  prices  is  but  little 


840  BUSINESS  CYCLES 

are,  it  is  impossible  to  speak  with  confidence  and  to  offer  anything  like 
proof.  We  are  inadequately  informed  about  the  facts  of  unemploy- 
ment in  this  country  until  the  unemployment  census  of  1930.1  From 
an  estimate  that  has  Professor  Wesley  C.  Mitchell's  sanction,2  no  "trend" 
can  be  deduced  that  would  have  any  meaning.  This  accords  with  pre- 
vious findings  (see  above,  I,  a,  3).  For  1920,  the  annual  figure  already 
displaying  the  influence  of  the  slump,  unemployment  was  5.1  per  cent 
of  nonagricultural  earners  of  wages  and  salaries.  The  slump  figures 
(1921  and  1922)  are  15.3  and  12.1  per  cent.  The  absolute  amount  corre- 
sponding to  the  former  percentage  (4.27  millions)  probably  "surpassed 
all  previous  records,"  but  there  is  no  reason  to  believe  that  the  percentage 
itself  did.  In  any  case  it  fell  back  to  5.2  per  cent  for  1923  and  rose  to 
7.7  in  1924.  The  years  1925  and  1926,  with  5.7  and  5.2,  indicate  an 
effect  of  the  Juglar  prosperity,  but  for  1927  we  have  again  6.3  per  cent. 
The  writer  thinks  that  the  figure  for  1928  was  higher  than  that  and  the 
figure  for  1929  certainly  not  lower  than  that  for  1928.  It  is  still  more 
difficult  to  interpret  these  figures  than  it  is  to  trust  them.  They  cer- 
tainly include  very  little  of  typically  spurious  or  "malingering"  unem- 
ployment, since  there  was  no  government  dole.  But  the  nomadic 
habits  of  the  American  workman  arid  the  high  level  of  earnings  that  made 
it  easy  to  tide  over  a  short  spell  of  unemployment  and  even  to  look  upon 
it  as  a  holiday,  suggest  that  those  figures  may  include  a  nonnegligible 
number  of  cases  that  bordered  on  voluntary  unemployment.  For  these 
and  other  reasons,  normal  unemployment  in  this  country  always  has 

weakened  by  the  knowledge  that  competitors  will  move  in  the  same  direction — the  stand- 
ard instance  is  once  more  provided  by  the  industry  that  typifies  entrepreneurial  behavior 
so  well,  the  motorcar  industry,  the  leading  firm  in  which  repeatedly  reduced  prices,  although 
it  must  have  known  that,  in  the  static  sense  adopted  by  Dr.  Sweezy  at  this  point,  this 
measure  would  be  self-defeating.  We  emphasize  this  because  so  much  has  been  made, 
in  the  discussion  of  these  problems,  of  special  cases  and  because  the  theory  of  imperfect 
competition  has  been  so  fertile  in  such  cases,  which,  though  interesting,  only  serve  to  con- 
fuse the  broad  issue. 

1  And  even  the  results  of  that  census  have  been  severely  questioned  by  competent 
critics. 

2  See  Recent  Economic  Changes,  1929,  II,  p.  879  and  the  chapter  on  labor.     The 
figures  are  intended  to  convey  an  idea  of  how  great  unemployment  has  at  least  been  in 
each  of  the  years  1920  to  1927.     If  it  actually  was  greater,  our  argument  in  this  and  the 
next  paragraph  would  apply  a  fortiori.     But  it  is  not  certain  that  it  was.     While  one  party 
to  the  discussion,  the  administration  included,  may  have  been  resolved  to  see  as  little 
unemployment  as  possible,  the  other  party  to  the  discussion  was  no  less  resolved  to  see 
as  much  as  possible.     The  tendency  to  exaggerate  is,  in  some  estimates,  as  obvious  as  is 
the  motive  for  it:  in  a  period  in  which  it  was  clear  that  the  majority  of  the  nonagrarian 
population  was  thoroughly  contented  with  the  results  of  the  capitalist  process,  the  critic 
had  little  else  to  fall  back  upon  and  naturally  made  the  most  of  it.     A  great  deal  depends 
upon  definition.     But,  however  we  define,  there  was  certainly  much  "unseen  unemploy- 
ment" in  addition. 


1919-1929  841 

been  higher  than  in  Europe.  If,  in  addition,  we  take  account  of  our 
expectation  from  the  prevailing  Kondratieff  phases,  the  conclusion  can 
only  be  very  tentative  that  an  estimate  of  minimum  unemployment 
which,  even  in  years  of  buoyant  activity,  never  fell  to  5  per  cent,  indi- 
cates the  presence  of  still  another  factor  that  made  for  unemployment. 
Direct  observation  of  everyday  practice,  however,  not  only  confirms 
this  conclusion  but  also  reveals  what  that  factor  was.  The  effort  to  be 
as  economical  as  possible  of  hired  labor,  which  has  been  described  above 
as  the  immediate  consequence  of  relative  dearness  of  labor,  was  in  fact 
an  obvious  feature  of  the  period  and  among  the  first  things  to  strike  a 
foreign  observer  of  American  industrial  and  private  life.  As  to  the 
former,  it  is  true  of  course  that  labor-saving  changes  in  methods  of  pro- 
duction would  have  come  about  in  the  normal  course  of  our  process  quite 
irrespective  of  the  level  of  wages.  But  many  of  them  were  conditioned 
by,  and  most  of  the  noninnovating  rearrangements  of  the  combinations  of 
factors  of  production  which  occurred  in  1920,  1921,  and  1922  were  clearly 
reactions  to,  a  price  of  labor  that  was  high  relatively  to  that  of  other  factors. l 
Money  cost  of  labor  per  unit  of  product  declined  substantially  to  1925 
and  then  again  from  1927  to  1929. 2  But  it  is  easy  to  see  that,  given  the 
high  long-run  elasticity  of  the  individual  firm's  demand  for  labor,  this 
fact  strengthens  the  case  instead  of  weakening  it.  Two  circumstances 
must,  however,  be  mentioned  which  worked  in  the  opposite  direction. 
First,  the  wage  rate  under  discussion  is  an  average.  Actual  rates  tended, 
on  the  whole  though  not  always,3  to  differ  from  each  other  in  a  manner 
that  would  mitigate  effects  on  employment:  they  were  as  a  rule  lower 
in  weaker  geographical  or  industrial  sectors — in  the  South,  for  instance, 
or  in  the  bituminous  coal  industry — and  higher  in  others  that  were  able 
to  bear  it.  Second,  a  wage  rate  high  enough  to  induce  substitution  of 
labor  by  other  factors — or  labor  of  the  "manual"  kind  by  other  labor — 
will  at  first  tend  to  increase  employment,  because  it  engenders  additional 
demand  for  labor-saving  devices,  most  of  which  have  themselves  to  be 
produced.  During  our  period  that  part  of  current  demand  for  labor- 

1  We  shall  observe  the  same  phenomenon  in  the  processes  of  recent  years.     To  the 
increases  in  wages  that  occurred  in  1936  and  1937,  for  instance,  industry  immediately 
reacted  by  "rationalization,"  which  in  many  cases  almost  succeeded  in  keeping  labor  cost 
per  unit  of  product  constant.     Southern  farmers,  not  immediately  affected  or  not  so  much 
affected  by  those  increases,  are  already  preparing  to  meet  them  by  mechanization. 

2  See  data  for  62  industries,  F.  C.  Mills,  Economic  Tendencies,  p.  404.     There  was  an 
interruption  of  the  decline,  which  in  those  data  shows  in  the  figures  for  1927,  that  accords 
with  our  expectation  for  a  Juglar  prosperity. 

3  Massachusetts,  or  even  the  whole  of  New  England,  is  a  conspicuous  geographical 
exception;  and  the  fate  of  her  industries  from  1923  to  1929  shaped  accordingly.     See  D.  H. 
Davenport  and  J.  J.  Croston,  Unemployment  and  Prospects  for  Reemployment  in  Massa- 
chusetts,  Publications  of  the  [Harvard]   Graduate  School  of  Business  Administration, 
Business  Research  Studies,  No.  15,  1936,  Chart  R  on  p.  63. 


842  BUSINESS  CYCLES 

saving  equipment  which  constituted  a  reaction  to  the  prevailing  wage 
level  no  doubt  helped  to  keep  up  total  demand  for  labor.  This  possible 
source  of  treacherous  "verifications"  of  high-wage  theories  should  always 
be  kept  in  mind  whenever  the  effects  of  wage  rates  on  employment  are 
being  discussed. 

That  the  price  of  labor  considered  as  a  consumers'  good  was  very 
high  relatively  to  the  prices  of  other  consumers 'goods  and  that  households 
reacted  accordingly1  is  too  obvious  to  detain  us.  It  should  merely  be 
added  that  this  reaction  went  even  further  than  appears  at  first  sight. 
The  mechanization  of  the  household  was  so  much  more  successful  in  this 
country  than  in  others  because  it  not  only  economized  labor  but  also 
made  it  easier  to  dispense  with  hired  labor  altogether.  Even  the  pros- 
perous American  family  thus  learned  to  perform  services  for  itself  which, 
but  for  the  rate  of  wages  and  especially  for  the  rate  of  wages  per  unit  of 
service,  it  would  have  delegated  to  nomnembers.  It  also  learned  to 
substitute  for  enjoyments  which  involve  the  direct  employment  of  labor 
others  which  do  so  less  or  not  at  all.  Thus  the  American  style  of  private 
life  was  powerfully  influenced,  dominated  perhaps,  by  the  level  of  wages.2 
Part  of  the  demand  for  durable  consumers'  goods  which  was  so  important 
a  feature  of  the  twenties  was  ultimately  nothing  else  but  a  flight  from 
labor  arid,  in  particular,  from  the  relatively  most  expensive  kind,  viz., 
manual  labor. 

It  might  be  asked  how  the  level  of  wages  can  have  influenced  employ- 
ment and  yet  not  have — appreciably — influenced  what  we  have  called 
the  general  complexion  of  business  situations.  The  answer  is,  first, 
that  the  high  wage  rates  themselves  largely  compensated  for  the  effects 
beyond  those  on  employment.  From  our  impression  that  more  labor 
would  have  been  employed  in  American  manufacturing  industry  if  wage 
rates  had  been  lower  than  they  were,  it  does  not  follow  that  its  pay  roll 
would  have  been  much  different  from,  still  less  that  it  would  necessarily 
have  been  higher  than,  what  it  actually  was  (see  below).  Second,  part 
of  the  additional  employment  which  would  have  accompanied  lower 
wage  rates,  would  not,  even  if  it  had  brought  about  a  substantial  net 

1  Households  also  reacted  directly  to  the  price  of  labor  in  its  productive  uses.     The 
relative  increase  in  the  use  of  consumers'  goods  requiring  less  labor  than  others,  and  the 
relative  decrease  in  the  use  of  consumers'  goods  requiring  more,  may  be  instanced  by 
the  shift  of  consumers'  demand  from  custom  tailored  to  ready-made  clothing. 

2  It  will  be  held  by  some  that  that  style  of  life  is  in  itself  an  achievement  in  efficiency 
and  by  others  that  it  spelled  increasing  democratization  and  a  moral  progress.     Certainly. 
The  writer  is  not  criticizing.     Also  it  might  be  thought  that  the  expansion  of  the  service 
industries  contradicts  the  statements  in  the  text.     But  a  moment's  reflection  should 
convince  the  reader  that  the  contrary  is  the  case,  for  certain  service  industries  are  nothing 
but  labor-saving  organizations  for  the  performance  of  services  which  without  them  would 
require  very  much  more  labor.    Apartment  hotels  "with  maid  and  valet  service"  are  an 
instance. 


1919-1929  843 

increase  in  the  total  of  the  national  wage  bill,  have  increased  total  expendi- 
ture on  commodities,  but  would  only  have  redistributed  it  among  con- 
sumers. This  is  best  seen  in  the  case  of  that  type  of  employment  which 
may  be  designated  by  the  term  help:  domestic  service,  shop  assistance 
in  small  retail  trade  and  in  the  business  of  the  carpenter,  painter,  plumber, 
and  so  on,  even  in  the  small  and  medium-sized  factories.  For  the  total 
amount  of  employment  these  cases,  persistently  neglected  by  the  current 
theories  of  wages,  are  extremely  important,  though  relatively  more  so 
in  Germany  and  England  than  in  the  United  States.  And  they  display 
a  particularly  simple  relation  between  employment  and  wage  rates. 
No  other  application  of  the  traditional  schema  of  economic  rationality 
is  so  obviously  true  to  real  life  as  the  picture  of  the  owner-manager 
of  a  retail  shop  who  "balances"  the  advantage  of  not  having  to  get  up 
early  in  order  to  sweep  his  shop  against  the  advantage  of  indulging  in 
marginal  glasses  of  beer.  But  although  changes  in  wage  rates  are 
promptly  reacted  to  in  those  fields  and  although  these  reactions  are  by 
no  means  indifferent  for  the  sum  total  of  satisfactions  and  incomes, 
they  do  not  affect  the  statistical  measures  of  industrial  output  or  the 
revenue  of  manufacturing  or  commercial  industry.  For  business  at 
large  it  is  indifferent  whether  the  retailer  hires  a  help  that  will  buy  con- 
sumers' goods,  or  does  the  spending  himself.  What  happens  in  most  of 
these  cases  is  that  some  earners  of  what  primarily  are  labor  incomes  share 
their  flow  of  consumers'  goods  with  others  in  return  for  the  latter's 
services.  Perhaps  another  million  men  and  women  could  have  been 
inserted  into  "productive"  activity  at  a  comparatively  small  sacrifice  in 
rates.  And  the  national  wage  bill  in  the  statistical  sense  might  have 
been  increased  correspondingly.  But  the  difference  this  would  have 
made  to  the  general  complexion  of  business  situations  would,  neverthe- 
less, have  been  small.1 

1  Those  of  us  who  believe  in  the  stronger  saving  propensities  of  the  shopkeeper  and  in 
their  catastrophic  effects  on  the  economic  process  will  have  to  rate  that  difference  more 
highly.  Even  so,  the  above  argument  would  hold  approximately.  There  would,  in  this 
case,  also  be  another  argument  against  high- wage  theories,  which  the  writer,  however, 
does  not  care  to  stress.  We  may  note  in  passing  still  another  line  of  reasoning  according 
to  which  a  higher  or  lower  level  of  wage  rates  acts  on  employment  and  output  through  the 
higher  or  lower  rates  of  interest  which  it  induces  not,  as  older  doctrine  would  have  argued, 
by  relatively  decreasing  or  increasing  the  supply  of  "real  capital,"  but  by  decreasing  or 
increasing  unused  lending  facilities.  It  is  true  that  variations  in  the  price  of  a  commodity 
quantitatively  so  important  as  labor  influence  all  other  prices  and  quantities  and  also  all 
monetary  magnitudes  and  expressions.  This  is,  in  fact,  why  a  complete  theory  of  wages  is 
so  very  complicated  a  matter.  But  the  simple  nexus  alluded  to  can  be  asserted  to  be 
operative  only  by  means  of  so  unrealistic  assumptions  and  owes  the  importance  attributed 
to  it  so  exclusively  to  a  theoretical  model  which  excludes  all  the  vital  mechanisms  through 
which  variations  in  wage  rates  act  that  we  need  not  proceed  with  it.  Under  its  assump- 
tions the  proposition  is,  of  course,  tautologically  true. 


844  BUSINESS  CYCLES 

It  will  be  seen  that  this  analysis  does  not  exclude  the  possibility  that 
still  higher  wage  rates  might  have  netted  a  higher  total  monetary  or 
even  real  wage  bill.  The  best  method  of  convincing  ourselves  of  this  is  to 
envisage  a  comprehensive  organization  of  all  employees  acting  as  a  dis- 
criminating monopolist.  This  monopolist  would  have  had  to  go  in  many 
spots  below  the  rates  that  were  actually  paid,  especially  if  he  had  acted 
with  a  view  to  maximizing  real  wage  bill  in  the  long  run.  But  that  both 
the  weighted  index  of  money-wage  rates  and  the  resulting  wage  bill 
might  have  been  higher  than  they  were  is  not  only  possible  but  plausible. 
Our  argument,  together  with  some  of  the  facts  glanced  at,  no  doubt 
suggests  that  in  that  case  total  output  might  have  been  smaller  or,  at 
all  events,  not  appreciably  greater  and  that  there  would  have  been  still 
more  unemployment.  But  all  we  have  established  is  that  employment 
of  labor  per  given  amount  of  product  and  direct  consumption  of  services 
of  labor  were  less  than  they  would  have  been  with  relatively  lower  wage 
rates.  Although  we  have  also  seen  that  this  component  of  total  employ- 
ment was  as  a  matter  of  fact  very  strong  and  that  any  increase  in  total 
output  which  might  have  been  produced  by  a  further  rise  in  wages  would 
have  had  to  be  very  great  to  counterbalance  it,  the  case  is  one  of  rea- 
sonable likelihood  only.  There  certainly  was  no  sign  of  "inadequate" 
consumers'  spending,  whether  due  to  saving  (admitting  for  argument's 
sake  that  saving  would  have  had  that  effect)  or  to  any  other  causes,  and 
there  were  reasons  more  convincing  than  that  for  such  underemploy- 
ment of  resources  as  there  was.  A  theory  of  the  world  crisis  can  be  no 
more  derived  from  any  effects  of  "too-low"  than  it  can  from  effects  of 
"too-high"  wages  and  all  the  attempts  in  this  direction — e.g.,  consumers' 
not  being  able  to  buy  what  was,  or  would  have  been,  produced — enter 
into  well-known  categories  of  provable  error.  But  that  does  not  consti- 
tute exact  proof  of  the  presence  or  absence  of  disturbance  emanating 
from  wage  rates.1 

3.  Much  of  the  above  applies,  mutatis  mutandis,  also  to  the  German 
case  (Chart  XLVIII). 

The  wage  bill,  owing  to  its  comprehensiveness,  carries  different 
meaning,  however,  and  the  data  about  unemployment  are  more  nearly 
exact,  although  it  is  still  necessary  to  allow  for  some  unseen  unemploy- 

1  The  reader  will  observe  that  this  is  exclusively  due  to  the  data  of  the  period  which, 
apart  from  their  inadequacies,  do  not  allow  us  to  insert  into  the  framework  of  general 
theory,  which  cannot  do  more  than  formulate  questions  and  describe  possible  cases,  factual 
assumptions  definite  enough  to  enable  us  to  choose  between  these  cases.  There  are 
other  situations — we  shall  meet  one  in  the  next  chapter — in  which  it  is  easier  to  do  this. 
Other  economists  feel  no  compunction  about  definite  assertions  in  any  case.  This  is  due 
partly  to  the  simplicity  of  the  theoretical  models  which  are  satisfactory  to  them,  partly 
to  the  readiness  to  embrace  particular  factual  assumptions. 


1919-1929 


845 


ment,1  and  although  a  change  in  the  attitude  of  the  public  mind  and  the 
possibilities  of  abuse  which  every  system  of  unemployment  insurance 
offers  may  have  had  something  to  do  with  the  absolute  figures.2  More- 
over, it  must  be  remembered — this,  of  course,  also  applies  to  the  United 
Kingdom — that,  while  in  the  United  States  hourly  earnings  were  (at 
least  they  would  be  if  we  imagine  them  to  be  calculated  with  ideal  cor- 


COST  OF  LIVING? 
WHOLESALE  PRICES 


1925 


1926 


1927         1928        1929         1930         1931          1932         1933 
CHART  XLVIII. — Germany  (see  Appendix,  p.  1071). 


1934        1935 


rectness)  approximately  equal  to  the  money  cost  per  hour  of  employing 
labor,  in  Germany  the  two  differed  significantly,  especially  by  employers' 
contributions  to  social  insurances. 

From  January  1925  to  December  1929  the  Reichsamt  figures  for 
average  hourly  trade-union  rates  (tarifliche  Stundenlohnsatze)  increased 
by  47  per  cent.3  This  increase  was  associated,  on  the  one  hand,  with  an 

1  By  unseen  unemployment  we  mean  unemployment  which  escapes  statistical  measure- 
ment, for  example,  because  it  outlasts  the  benefit  period.     We  do  not  mean  unemployment 
which,  being  only  the  result  of  artificial  definitions,  exists  but  in  the  pages  of  some  theorists 
and  may  include  individuals  who  think  themselves  fully  employed. 

2  There  is  no  doubt  that  those  abuses  have  been  recklessly  exaggerated  by  some  and 
recklessly  denied  by  others.     The  war  waged  by  the  authorities  upon  what  was  termed 
black  labor  indicates,  however,  that  they  were  not  negligible.     But  we  need  not  believe, 
or  attach  much  importance  to,  the  drastic  anecdotes  that  were  current  about  them. 

8  They  continued  to  increase  in  1930,  but  not  much.  It  must  be  noticed,  however,  that 
the  "effective"  rates  (=  official  basic  rates  plus  various  additions,  Zuschltige),  which- we 


846  BUSINESS  CYCLES 

increase  of  about  10  per  cent  in  cost  of  living,  and  on  the  other  hand, 
with  a  substantial  increase  in  unemployment.1  Neither  was  simply 
"due"  to  that  increase  in  rates.  Concerning  the  first,  there  were  several 
other  factors,  the  most  important  of  which  were  the  agrarian  policy 
(effective  in  the  last  3  years  of  the  period)  and  the  gradual  normaliza- 
tion of  house  rent  (expenditure  on  shelter  increased  by  55  per  cent  from 
1925  to  1929).  But  there  remains  a  residue  which,  being  contrary  to 
Kondratieff  expectation,  can  hardly  be  attributed  to  anything  else  but 
to  increase  in  wage  rates.  This  would  mean,  of  course,  that  the  latter 
was  effective  in  raising  the  monetary  wage  bill  above  what  it  would 
otherwise  have  been:  for  that  increase  in  wage  bill  which  is  the  "natural" 
consequence  of  the  downgrade  process  would  not,  as  we  know  from  both 
theory  and  history,  increase  the  cost  of  living.  It  is,  however,  perfectly 
plausible  that  in  this  case  and  for  the  time  being  the  increase  in  rates 
actually  resulted  in  a  considerable  increase  in  wage  bill,  because  it  was 
crowded  into  less  than  4  years,  which  is  hardly  enough  for  the  system  to 
adapt  itself  fully. 

Concerning  unemployment,  the  relevance,  not  only  of  burdens  directly 
associated  with  the  employment  of  labor,  but  also  of  burdens  apparently 
unconnected  with  it,  must  again  be  stressed.  We  have  seen  that  high 
corporation  and  income  taxes  increase  the  sensitiveness  of  the  economic 
process  to  most  other  disturbances  and  in  particular  to  any  increase  in 
costs.  Since  taxation  of  this  type  was  the  outstanding  economic  feature 
of  postwar  Germany  and  higher  than  anywhere  else,  it  is  not  possible  to 
disentangle  the  effects  which  the  increase  in  wage  rates  might  have  had 
if  it  had  occurred  alone.  Perhaps  there  was  some  truth  in  the  contention 
of  the  exponents  of  trade-union  interests  that  the  wage  rates  did  not 
per  se,  at  least  temporarily,  substantially  contribute  to  the  abnormal 
and  rising  amount  of  unemployment  that  prevailed  practically  through- 
out.2 We  can  certainly  not  use  the  reasoning  that  served  in  the  American 
case,  for  German  wage  rates  started  from  what  even  in  1925  was  a  low 
level.  But  what  the  high  general  level  kept  by  wage  rates  throughout 

have  for  only  some  industries,  rose  less.  Even  if  we  take  "tariff"  rates,  their  increase 
does  not  mean,  of  course,  that  even  in  1930  they — or  earnings — were  high  according  to 
general  cultural  standards.  Real  hourly  rates  were  only  about  10  per  cent  above  1914  in 
1929  and  only  16  per  cent  in  1930.  On  the  other  hand,  if  it  is  cultural  standards  and  welfare 
considerations  that  we  have  in  mind,  the  considerable  unearned  increment  in  real  incomes 
must  be  taken  into  account  which  accrued  to  the  working  class  from  public  expenditure. 

1  The  hyperseasonal  peak  in  the  winter  1928-1929  is,  however,  due  to  the  indirect 
effects  of  the  labor  struggle  in  the  Ruhr  district  (the  direct  effects  of  strikes  and  lockouts 
are  excluded).     The  peak  in  the  winter  1923-1924,  when  about  one-fourth  of  all  members 
of  trade  unions  was  unemployed,  of  course,  does  not  count  either  for  our  purpose. 

2  The  long  and  acrimonious  controversy  which  was  at  the  time  waged  on  the  subject 
has  lost  its  interest  because  the  arguments  used  by  either  party  sound  hopelessly  antiquated 
now.     Even  if  they  be  appraised  ex  visu  of  their  dates,  little  can  be  said  for  most  of  them. 


1919-1929  847 

the  period  did  in  this  country,  was  in  Germany  largely  done,  though  by 
a  different  route,  by  the  high  general  level  of  taxation:  wage  rates, 
"social  burdens,"  and  taxes  taken  together  may  reasonably  be  held  to 
explain  conditions  in  the  labor  market  which  neither  of  them  could  explain 
if  taken  in  isolation.1 

4.  This  is  not  less  true  for  the  United  Kingdom.  Otherwise,  the 
case  is,  at  least  in  one  important  respect,  more  like  the  American  than 
like  the  German.  We  find  remarkable  stability  of  money  wage  rates 
from  the  end  of  1924  to  the  beginning  of  1928  and  then  a  slight  decline, 
which  was  to  become  somewhat  more  pronounced  in  1930.  As  in  the 
United  States,  this  was  preceded  by  a  downward  revision  from  the  post- 
war peak,  which  extended  from  January  1921  to  the  end  of  1923,  and 
by  a  recovery  from  that  which  covered  19242  but  was,  in  contrast  to 
what  happened  in  America,  only  slight.  Nevertheless,  it  left  money 
wages  at  about  96  per  cent  above  their  1913  figure — already  associated 
with  some  supernormal  unemployment — a  level  which  had  only  to  be 
kept  up  in  order  to  increase  real  wages,  thanks  to  the  Kondratieff  ten- 
dency, free  trade,  and  the  monetary  policy  pursued,  to  nearly  20  per 
cent  above  1913  at  the  end  of  our  period  (annual  average).  See  Chart 
XLIX. 

This  is  what  trade  unions  and  the  various  public  agencies3  attempted 
and,  during  our  period,  achieved.  The  slight  upward  pull  exerted  by 
the  Juglar  prosperity  is  not  visible  in  the  chart  and  barely  so  in  the 
series — the  index  increased  by  one  point  for  a  short  spell  (December 

1  The  above  is  confined  to  the  years  which  followed  upon  inflation,  stabilization  and  its 
immediate  effects.     As  mentioned  in  a  previous  note,  there  was  "disturbance  unemploy- 
ment" in  the  winter  1923-19124.     But  it  is  worth  noting  that  the  rest  of  1924  displayed 
only  a  moderate  number  of  totally  unemployed — a  fact  which  was  associated  with  a  very 
moderate  level  of  rates. 

2  There  is  a  statistical  difficulty  about  these  statements.     We  are  using  Professor 
Bowley's  index  of  weekly  wages  (see  Appendix,  description  of  Chart  XLIX),  which  from 
January  1925  runs  on  a  new  basis.     The  two  series  have  been  "spliced/'  but  no  great 
confidence  can  be  placed  in  this  procedure.     This  is  why  we  refrain  from  statements  about 
the  course  of  wages  during  the  war.     According  to  the  old  series  (not  shown  in  the  chart, 
except  as  transformed  by  splicing),  rates  would  have  followed  cost  of  living  with  a  lag  and 
caught  up  with  it  in  1919,  more  definitely  in  1920.     The  maximum  of  rates  (January  1921 ; 
277  per  cent  of  1913)  would  have  followed  the  maximum  of  cost  of  living  (278  per  cent) 
with  a  lag  of  three  months. 

3  Along  with  the  increase  in  the  power  of  trade  unions,  public  regulation  of  wages  devel- 
oped from  the  Minimum  Wage  Act  of  1909  to  the  Amending  Act  of  1918,  the  Joint  Indus- 
trial Councils,  the  Railway  Conciliation  Councils,  and  the  Agricultural  Wage  Boards.     If 
we  include  the  parallel  activities  of  trade  unions  and  take  account  of  the  fact  that  official 
awards  will  exert  influence  beyond  the  cases  decided  and  beyond  the  trades  under  jurisdic- 
tion, we  may  say  that  almost  all  wage  contracts  in  the  United  Kingdom  are  publicly 
controlled.     This,  of  course,  raises  a  problem  in  the  interpretation  of  our  curves.     But  it  is 
believed  that  what  follows  in  the  text  is  not  open  to  objection  on  this  ground. 


848 


BUSINESS  CYCLES 


1926  to  April  1927) — and  the  unusual  lack  of  covariation  with  profits 
which  have  been  inserted  in  this  chart  in  order  to  display  it,  is  significant. 
These  features  as  well  as  the  attendant  unemployment,  no  doubt,  invite 
an  interpretation  similar  to  that  we  have  tentatively  adopted  in  the 
American  case.  This — i.e.,  that  the  general  level  of  wages  was  "too 
high"  from  the  start — was,  in  fact,  the  opinion  of  most  English  econ- 
omists,1 and  an  impression  to  that  effect  must  have  prevailed  among 


1919     1920    1921     1922    1923    1924    1925    1926    1927    1928    1929    1930    1931     1932    1933    1934    1935 
CHART  XLIX.— United  Kingdom  (see  Appendix,  p.  1072). 

trade-union  leaders,  who,  though  they  energetically  defended  that  level, 
never  seriously  tried  to  raise  it.  The  present  writer  doubts  its  correct- 
ness, however.  That  increase  in  real  wages  was  not  so  very  impressive. 
It  was  certainly  smaller  than  it  would  have  been  without  the  war  and 
hardly  greater  than  we  should  expect  it  to  be  in  spite  of  the  war,  during  a 
Kondratieff  downgrade  and  in  a  country  that  as  yet  substantially 
adhered  to  free  trade.  Any  explanation  of  strains  and  maladjustments 
that  rests  on  an  increase  of  real  wages  by,  from  1924  to  1929,  little  more 

1  That  fact  is  veiled  not  only  by  the  guardedness  of  the  statements  of  some  of  them 
but  also  by  the  fact  that  the  concurrence  of  others  was  expressed  in  an  indirect  manner, 
which  is  likely  to  escape  notice.  Advocacy  of  "monetary  expansion"  may,  advocacy  of 
protective  tariffs  must,  imply  attack  on  real  wages  while  it  is  at  the  same  time  natural  for 
the  advocate  of  these  measures  to  oppose  a  reduction  in  money  wages. 


1919-1929  849 

than  1  per  cent  per  year  seems  hazardous,  if  not  downright  misleading.1 
The  national  wage  bill  became  burdensome  because  it  was  accompanied 
by  a  fiscal  policy  that  made  it  so.  If  the  labor  interest  did  not  press  for 
an  increase  in  wage  rates,  it  attained  the  same  object  by  pressing  for, 
or  putting  out  of  court  the  reform  of,  taxation  which  effected  a  transfer  of 
wealth2  much  greater  than  any  conceivable  increase  in  wage  rates  could 
have  effected. 

Many  other  factors,  however,  swelled  unemployment  percentages 
during  the  period  under  discussion.  The  great  problem  that  attracted 
so  much  anxious  attention  dates  from  1923.  Until  the  last  quarter  of 
1920  the  trade-union  percentage  (unemployed  males)  was  understand- 
ably below  anything  that  can  be  called  normal  according  to  any  stand- 
ard. It  soared  to  over  23  per  cent  in  June  1921,  but  this,  too,  was  readily 
understandable.  In  1923,  however — we  now  shift  to  the  percentage  of 
unemployed  insured  persons,  which  was  (annual  average)  17  in  1921  and 
14.3  in  1922 — unemployment  was  still  at  11.7  per  cent  and  in  the  vicinity 
of  that  figure  it  stayed3  throughout  (1924,  10.3;  1925,  11.3;  1926,  in  spite 
of  strikers  disqualified  for  benefit  being  excluded,4  12.5;  1927,  9.7;  1928, 
10.8;  1929,  10.4).  There  is,  as  has  been  stated  before,  no  significant 
"trend"  in  this  but,  even  apart  from  the  figure  for  1926,  which  reflects 
the  indirect  influences  of  the  great  struggle  of  that  year,  a  "level"  of 
about  10  per  cent.  Now  this  cannot  be  compared  to  prewar  (trade-union) 
percentages,  because  the  percentage  of  total  unemployment  is,  of  course, 
likely  to  be  greater  than  the  trade-union  percentage  which  served  as 
indicator  in  the  prewar  epoch;  because,  for  the  same  reasons  as  in  Ger- 
many, statistically  visible  unemployment  would  presumably  be  greater 
in  the  postwar  period,  even  if  we  had  otherwise  comparable  figures; 
and  because  trade-union  regulation,  social  insurance,  and  other  factors 
made  for  reduced  geographical  and  industrial  mobility  of  labor.  For 
these  reasons  normal  unemployment  in  our  sense  of  the  term  would, 
ceteris  paribus,  be  greater  than  it  had  been.  It  has  been  estimated  that 
even  in  times  of  active  business  it  would  not  fall  below  from  2  to  4  per 
cent,5  and  we  will  tentatively  accept  3  per  cent  as  a  compromise.  Second, 

1  The  above  statement  is  intended  to  be  read  with  previous  discussions  in  mind  that 
cannot  be  repeated.     Unconnected  with  these,  it  would,  of  course,  be  meaningless.     It 
should  be  added  that  that  increase  in  real  wages  was  largely  at  the  expense  of  foreign 
producers. 

2  The  reader  will  observe  that  the  above  statement  stresses  taxation  only  and  not  also 
expenditure  benefiting  the  masses.     For  such  as  it  actually  was,  this  would  have  been 
possible  with  almost  no  interference  with  the  efficiency  of  the  capitalist  machine. 

3  Trade-union  figures  were,  however,  materially  lower  in  1924  (minimum  of  7  per  cent 
in  May). 

4  So  they  were  in  1921. 

*  See  Colin  Clark,  Statistical  Studies,  Economic  Journal  for  September  1931,  p.  849. 
The  trade-union  percentage  was,  however,  as  low  as  0.9  in  April  1920. 


850  BUSINESS  CYCLES 

there  is  the  normal  effect  of  the  Kondratieff  downgrade  to  take  account 
of,  which  may,  according  to  the  experience  of  the  seventies  and  eighties, 
easily  double  that  figure.  This  already  includes  the  industrial  and 
commercial  shifts  and  readjustments  incident  to  the  downgrade  processes, 
but  only  so  much  of  them  as  the  internal  evolution  of  the  country  would 
entail.  The  great  shift  from  world  market  to  home  production  which 
English  industry  had  to  undergo  in  consequence  of  the  shrinkage  of  her 
exports  is,  third,  not  covered  by  that  estimate.  If  we  accept  Mr.  Colin 
Clark's  figure1  that,  if  England's  "position  as  an  exporting  co*untry  had 
not  been  deteriorating,  unemployment  in  1929  would  have  averaged 
about  900,000  as  against  1,250,000,"  and  neglect  the  possibility  that 
this  deterioration  may  also  have  had  something  to  do  with  wages,  we 
have  here  an  independent  factor  the  influence  of  which  we  will  put  at 
2  per  cent  in  the  average,  which  is  certainly  conservative.  Thus  we 
arrive  at  the  result  that  8  out  of  the  10  per  cent  unemployment  level  of 
the  period  can  plausibly  be  accounted  for,  if  not  entirely  without  reference 
to  labor  policy,  yet  without  reference  to  any  behavior  of  wage  rates  other 
than  what  it  would  have  been  within  the  most  normal  working  of  our 
process  and,  substantially,  even  under  laissez-faire  conditions.  This 
statement  is  open  to  objection  on  statistical  and  theoretical  grounds;2 
but  it  still  serves,  so  the  writer  believes,  to  give  a  rough  idea  of  the  order 
of  magnitude  of  what  may  remain  for  explanation — as  vicarious  unem- 
ployment— by  "rigid"  wage  rates  plus  fiscal  policy.  However  that  may 
be,  it  cannot  well  be  doubted  that  the  postwar  unemployment  problem 
was  largely  one,  not  of  systematic  tendencies,  but  of  disturbance  by 
external  factors. 

III.  In  proceeding  to  discuss  the  course  of  events  in  the  banking  and 
cognate  spheres  we  will  remind  ourselves  again  that  this  sequence  of 
topics  is  not  to  suggest  a  progress  from  effects  to  causes  and  that,  though 
there  is  nothing  but  interdependence  between  the  quantities,  monetary 
and  other,  which  enter  into  our  process,  we  should,  were  we  constrained 
to  set  up  causal  nexus  at  all,  prefer  to  put  our  trust  in  the  reverse  one. 

1  Op.  cit.,  p.  349.     See  ante,  Sec.  E,  1. 

2  The  indubitable  truth,  stressed  by  Professor  Pigou  in  his  Theory  of  Unemployment, 
that  it  is  impossible  to  distribute  a  given  amount  of  unemployment  among  different  causes, 
does  not,  however,  stand  in  our  way,  because  there  is  no  objection  to  trying  to  estimate 
the  difference  which  presence  or  absence  of  a  given  factor  makes  or  would  make  and  it  is 
in  this  sense  that  our  statement  should  be  understood.     But  it  is  more  serious  that  the 
other  factors  would  of  course  have  produced  different  results  at  different  wage  levels  and 
can  never  be  treated  independently  of  them.     We  restate,  therefore:  Assuming  that,  given 
the  fiscal  and  monetary  policies  actually  pursued,  the  policy  of  trade  unions  (including  the 
threat  of  general  unrest)  and  of  the  public  wage-fixing  agencies  kept  money  wage  rates  above 
what  they  would  have  been  without  those  unions  and  agencies,  then,  whatever  that  dif- 
ference was,  its  responsibility  for  unemployment  can  at  most  have  amounted  to  2  per  cent 
of  the  total  of  insured  persons. 


1919-1929  851 

This  reminder  is  justified  by  the  fact  that  it  was  primarily  with  reference 
to  postwar  developments  and  in  connection  with  postwar  problems  that 
the  "deposit  logic"  committed  its  most  flagrant  excesses.  It  should 
also  be  borne  in  mind  that  in  no  country,  but  especially  not  in  the  United 
States,  did  banking  figures  retain  their  prewar  significance — not  so  much 
because  of  statistical  reasons  as  because  of  fundamental  institutional 
changes:  all  the  more  important  is  it  to  see  the  old  essence  under  new 
forms  and  phraseologies  where  the  old  essence  did  persist.1  Our  dis- 
cussion will  almost  wholly  be  confined  to  the  American  developments. 
a.  We  begin  by  clearing  up  a  point  which  may  have  puzzled  the 
reader  in  inspecting  Chart  XL VI.  While  in  the  United  States  postwar 
pulse  chart  balances  have  been  represented,  in  deference  to  prevailing 
opinion,  by  outside  net  demand  deposits  plus  "circulation"  (for  explana- 
tion, see  Appendix),  debits  have  been  compared  on  Chart  XL VI  with 
total  outside  deposits  (demand  plus  time)  minus  outside  investments. 
According  to  that  prevailing  opinion,  only  demand  deposits  are  balances 
in  our  sense  or,  as  most  authors  prefer  to  put  it,  "money."2  We  have  a 
good  series  for  what  is  referred  to  as  adjusted  demand  deposits,  which  is 
cleared  of  interbank  and  includes  government  deposits,3  and  most  of  us 
use  net  demand  deposits  as  a  substitute  for  purposes  for  which  the  other 
series  is  not  available.  But  time  deposits  are  held  to  have  no  more  claim 
to  be  included  than  bonds  would  have  were  banks  in  the  habit  of  issuing 
them.  This  view  seems  correct  as  far  as  time  deposits  represent  house- 
hold investments — genuine  "saving  deposits" — but  for  the  following 
reasons  it  does  not  seem  to  be  so  with  respect  to  the  bulk  of  time  deposits 
in  commercial  banks — and  a  great  part  of  the  time  deposits  in  savings 
institutions — which  does  not  constitute  investment  and  has  nothing  to 
do  with  savings  and  the  spectacular  increase  in  which  it  is  a  mistake  to 
treat  as  an  indication  of  changes  in  the  rate  of  saving.4 

1  It  may  also  be  mentioned  that  the  figures  of  national  banks,  notwithstanding  the 
changes  made  in  their  position  by  the  Federal  Reserve  Act  and  the  fact  that  their  propor- 
tionate importance  has  been  on  the  decrease,  might  still  be  used  for  our  period  with  approxi- 
mately as  much  justification  as  there  was  in  the  prewar  time.     The  late  A.  A.  Young  was 
of  the  same  opinion.     We  shall,  with  one  exception,  not  go  into  seasonal  variations,  which 
have  been  considerably  affected  by  the  institutional  changes,  and  will  hence  mention  here 
that  the  autumnal  drain  of  cash  from  New  York  had  lost  much  of  its  importance.     On 
seasonal  variations  in  general,  see  A.  A.  Young,  Analysis  of  Bank  Statistics,  p.  53  et  aeq. 

2  We  do  not  now  follow  that  usance  in  order  to  avoid  a  needless  controversy.     The 
term  Means  of  Payment  is  less  open  to  objection,  but  may  still  include  things  that  nobody 
wishes  to  include,  owing  to  the  ambiguity  of  the  word  payment. 

3  See  L.  Currie,  The  Supply  and  Control  of  Money,  1934,  p.  13.     Study  of  Chap.  Ill 
of  that  book  is  recommended  as  an  introduction  into  the  statistical  difficulties  of  the 
subject. 

4  The  exaggerated  ideas  some  writers  entertain  about  postwar  saving  activity  are 
sometimes  precisely  due  to  that  mistake.     It  has  been  pointed  out  before,  for  example, 


852  BUSINESS  CYCLES 

First,  as  far  as  time  deposits  were  actually  drawn  against,  they  served 
from  the  standpoint  of  holders  exactly  the  same  purpose  as  demand 
deposits.1  Opinions  differ  as  to  the  extent  of  the  practice  (which  was 
prohibited  in  1933),  but  its  importance  cannot  be  read  off  from  statistics 
of  turnover.  This,  of  course,  is  much  smaller  than  that  of  demand 
deposits.  But  on  the  one  hand,  it  is  naturally  slow-moving  "cash"  that 
is  held  on  time  deposits  and,  on  the  other  hand,  it  is  the  possibility  of 
drawing  (practically)2  at  will  that  matters.  Second,  even  if  no  check 
had  ever  been  drawn  against  time  deposits  and  if,  as  in  the  case*  of  bonds, 
one  would  really  have  had  to  "convert  them  into  money,"  they  would 
still  have  been  so  very  like  cash — since,  unlike  bonds,  they  can  always  be 
turned  into  cash  at  par — that  the  making  of  the  distinction  comes,  from 
the  standpoint  of  the  holder,  very  near  to  hairsplitting.  Whoever  holds 
that  kind  of  asset  will  behave  differently  with  respect  to  his  demand 
balances,  in  particular,  feel  much  less  constrained  than  he  otherwise 
would  to  keep  an  emergency  reserve.  By  the  classing  of  time  deposits 
with  customers'  investments  an  important  feature  of  the  monetary  situ- 
ation is  completely  lost.3 

Third,  a  growing  habit  of  keeping  on  time  account  as  much  of  one's 
cash  balance  as  is  possible  or  convenient  would  suffice  to  explain  the 
growth  of  time  deposits  relatively  to  demand  deposits.  Suppose  that 

that  Mr.  Lough's  estimates  of  savings  are  partly  due  to  his  treatment  of  time  deposits, 
although  he  is  not  unaware  of  the  point  and  tries  to  meet  it. 

1  Dr.  Currie,  op.  cit.t  p.  15,  argues  ably  against  this  view.     But  even  if  the  present 
writer  had  more  confidence  in  the  realistic  virtues  of  Dr.  Currie' s  assumptions,  that  argu- 
ment would  not  meet  the  point  as  formulated  above. 

2  A  bank  that  refused  to  honor  a  check  covered  by  a  time  deposit,  had  to  be  prepared  to 
face  an  unpleasant  discussion  and  to  lose  the  customer.     The  writer  in  some  cases  observed 
that  clerical  staffs  had  blanket  power  to  honor  such  checks  but  had  to  refer  to  a  responsible 
executive  if  they  wanted  to  refuse.     Of  decisive  importance,  finally,  are  those  agreements 
that  seem  to  have  been  fairly  frequent,  especially  in  the  West — so  the  writer  has  been 
told  by  a  competent  authority — according  to  which  holders  of  time  deposits  were  accorded 
a  limited  right  to  draw  without  notice,  two  or  three  times  a  year  for  example.     This  is 
proof  that  time  deposits  were  by  those  banks  looked  upon  as  a  special  kind  of  demand 
deposit  to  which  time-deposit  privileges  in  the  shape  of  higher  interest  were  granted  under 
the  pressure  of  competition.     The  Report  of  the  Committee  on  Member  Bank  Reserves, 
1931,  seems  to  take  much  the  same  view.     Mention  is  due  to  Dr.  B.  Anderson's  important 
contributions  to  the  subject  of  time  deposits,  see,  in  particular,  Bank  Expansion  versus 
Savings,  Chase  Economic  Bulletin,  June  25,  1928. 

8  It  might  be  urged  that  a  similar  consideration  applies  to  all  highly  liquid  assets.  This 
is  quite  true  to  the  extent  that  neglect  of  any  type  of  what  we  have  called  "near-money" 
is,  in  fact,  a  bar  to  correct  analysis  of  monetary  processes  and  a  cause  of  the  inadequacy  of 
the  picture  drawn  by,  as  well  as  of  the  recommendations  of,  the  modern  exponents  of  the 
quantity  theory  or,  to  use  Dr.  Anderson's  phrase,  those  economists  who  know  nothing 
but  "the  monotonous  tit-tat-to — money,  credit,  prices."  But  we  might  reply  that  such 
assets  cannot  be  normally  relied  on  to  represent  so  prompt  a  control  over  a  definite  sum 
as  do  time  deposits. 


1919-1929  853 

there  are  in  a  country  only  demand  and  "genuine  savings  deposits/* 
The  former  all  "circulate"  but  display  different  rates  of  activity  ranging 
from  several  hundred  turnovers  to  almost  zero  per  year.  Then  let  a  new 
name,  say  time  accounts,  and  some  privilege  be  introduced  for  balances 
which  are  being  used  less  often  than  an  arbitrary  number  of  times  per 
arbitrary  period,  people  being  invited  to  register  for  these  time  accounts 
on  condition  that  they  use  them  only  at  certain  dates  or  at  a  certain 
notice.  In  the  name  of  all  that  we  know  about  the  behavior  of  people, 
we  must  assume  that  they  actually  will  register,  first  and  foremost,  with 
respect  to  that  part  of  their  demand  deposits  which  they  would  anyhow 
not  use  before  the  assigned  dates  or  need  not  use  without  notice,  and 
second,  with  respect  to  another  part  which  can  be  made  to  qualify  with 
but  little  inconvenience.  Nothing  whatever  has  changed  in  the  behavior 
of  those  who  register.  In  particular,  the  holders  of  these  new  time  accounts 
spend  exactly  or  almost  exactly  what  and  when  they  used  to  spend  before. 
Yet  time  deposits  have  emerged  which,  incidentally,  may  reflect  credit 
creation  just  as  much  as  demand  deposits. 

This  paradigma  is  merely  intended  to  illustrate  a  principle.  It  is 
not  held  that  the  new  practice  will  not  make  any  difference  in  other 
respects:  if,  in  particular,  legislation  imposes  higher  reserve  proportions 
on  demand  than  on  time  deposits,  the  shift  of  balances  from  the  former 
into  the  latter  category  will  increase  the  lending  power  of  banks,  which 
is  precisely  how  the  growth  during  the  twenties  of  time  deposits  in  this 
country  became  an  instrument  of  monetary  expansion;  and  any  payment 
out  of  time  deposits,  no  matter  whether  effected  directly  or  by  means  of 
a  previous  conversion  into  demand  deposits,  will,  if  it  induces  a  more 
than  momentary  shift  from  time  to  demand  account,  have  the  opposite 
effect,1  the  same  as  an  increase  in  reserve  requirements  would  have. 

1  Dr.  Currie  rightly  emphasizes  this.  If  the  bank  in  question  happened  to  be  "loaned 
up,"  it  would,  in  consequence  of  such  a  payment,  have  to  reduce  its  loans  or  investments 
by  a  corresponding  though  not  equal  amount.  But  if  time  deposits  are  freely  flowing  in 
or  demand  deposits  have  a  tendency  to  shift  to  time  account — an  essentially  temporary 
state  of  things,  no  doubt,  but  one  which  prevailed  during  the  twenties — the  bank  will 
in  the  typical  case  be  relieved  of  that  necessity,  even  if  it  were  so  perfectly  loaned  up  as 
Dr.  Currie  must,  by  virtue  of  his  theoretical  scheme,  assume  any  bank  on  principle  to  be 
and  even  if  hence  any  use  made  of  time  deposits  would  necessarily  entail  some  contraction 
in  other  demand  deposits.  Perhaps  another  comment  will  be  useful,  though  it  is  really 
implied  in  what  has  been  said  in  the  text.  Let  us  return  to  the  hypothesis  that  payment 
can  be  effected  only  by  draft  on  demand  deposits  so  that  time  deposits  can  serve  the 
purpose  in  no  other  than  the  indirect  way,  i.e.,  through  previous  acquisition  of  a  demand 
deposit.  Then  firms  and  households  will  for  each  day  acquire,  against  time  deposits, 
demand  deposits  sufficient  to  provide  for  the  expected  payments  of  that  day.  But  not 
more.  The  rest  of  the  balance  will  stay  on  time  account.  Demand  deposits  and  bank 
reserves  will  be  "economized"  thereby.  But  exactly  as,  ex  visu  of  any  given  day,  we  do 
not  confine  the  concept  of  deposits  to  that  amount  of  demand  deposits  which  changes 


854  BUSINESS  CYCLES 

This  is  why  we  emphasized  the  growth  or  spread  of  the  habit  to  insure 
that  there  should  always  be  a  net  shift  toward  time  account.  We  may 
further  approximate  the  paradigma  to  American  reality  by  granting 
that  many  big  concerns  and  rich  households  did  not  stoop  to  availing 
themselves  of  the  advantages,  in  their  nature  modest,  which  we  suppose 
to  have  been  attached  to  time  accounts,1  so  that  it  may  be  primarily  the 
middle  and  lower  strata  of  depositors  who  did  so;  by  allowing  for  savings 
institutions'  trying  to  acquire  slow-moving  balances,  once  the  trick  has 
been  learned,  while  commercial  banks  had  an  additional  motive  to  hunt 
for  genuine  saving  deposits,  as  soon  as  they  were  in  turn  privileged  in 
handling  them;2  and  finally,  by  adding  that  temporarily  idle  funds  will 
naturally  tend  to  be  placed  on  time  account.3 

hands  on  that  day  but  extend  it  for  obvious  reasons  to  the  total  of  demand  deposits  in 
existence,  so  we  would  for  the  same  reasons  have  to  include  the  time  deposits  that  are  not 
converted  into  demand  deposits  and  hence  remain  time  deposits,  even  if  that  part  which 
is  converted  were  always,  necessarily  and  entirely,  balanced  by  a  contingent  contraction 
of  other  demand  deposits.  This  is  what  our  paradigma  is  intended  to  show.  The  impor- 
tant but  nonessential  difference  in  reserve  requirements  has  certain  consequences,  but  it 
does  not  alter  the  nature  of  time  deposits. 

1  Statements  about  banking  practice  differ  on  this  point.     Dr.  Anderson  holds  that 
the  time  deposits  of  New  York  City  banks  chiefly  consisted  "in  temporarily  idle  funds  of 
great  business  corporations  or  of  foreign  banks  or  of  rich  investors  who  had  temporarily 
disposed  of  investments  and  were  awaiting  opportunities  in  the  market."      (A  Critical 
Analysis  of  the  Book  by  Lauchlin  Currie,  address  before  the  New  York  chapter  of  the 
American  Statistical  Association,  Apr.  26,  1935,  p.  20)  while  others  hold  that  "bank 
statistics  show  clearly  that  there  could  have  been  no  appreciable  amount  of  shifting  of 
deposits  from  the  demand  to  the  time  category  on  the  part  of  large  depositors  in  national 
banks"  (D.  R.  French,  Significance  of  Time  Deposits  in  the  Expansion  of  Credit,  1922-1928, 
Journal  of  Political  Economy  for  December  1931).     The  two  statements  do  not  conflict 
as  much  as  they  seem  to.     But  it  is  not  unlikely  that  those  big  concerns  which  had  such 
good  opportunity  for  temporary  investment  resorted  less  to  time  accounts  than  did  smaller 
ones  that  had  not.     And  it  is  just  as  well  to  note  that  this  point  is  not  essential  to  our 
argument. 

2  Of  course,  both  savings  institutions  and  commercial  banks  always  did  do  that,  as  we 
have  seen  before,  in  this  country  as  well  as  in  others.     The  Federal  Reserve  Act  only  added 
another  stimulus  and,  for  national  banks,  additional  powers. 

3  We  hold,  agreeing  in  this  with  Dr.  Anderson,  that  the  bulk  of  time  deposits  has  much 
more  to  do  with  credit  creation  than  with  savings.     This  is  not  refuted  by  the  fact  that 
deposits  evidenced  by  passbooks  averaged  almost  72  per  cent  of  total  time  deposits  in 
national  banks  since  the  autumn  of  1928,  from  which  Mr.  D.  M.  Dailey  (National  Banks 
in  the  Savings  Deposits  Field,  Journal  of  Business  for  January  1931)  infers  that  "in  excess 
of  four-fifths  of  the  time  deposits  reported  by  the  national  banks  represent  the  accumula- 
tion of  savings:"  there  is  no  reason  for  inferring  that,  though  there  is  for  taking  the  large 
use  made  of  the  passbook  method  as  an  indication  that  it  was,  in  fact,  firms  and  households 
of  moderate  means  which  primarily  availed  themselves  of  the  opportunity  offered  by  time 
accounts.     But  while  denying  the  alleged  relation  of  time  deposits  to  saving,  we  do  not 
entirely  deny  their  relation  to  underspending.     In  order  to  bring  out  what  seems  to  the 
writer  an  important  principle,  our  paradigma  has  been  so  framed  as  to  be  independent  of 


1919-1929  855 

Fourth,  a  motivating  privilege  such  as  has  just  been  envisaged,  was 
provided  by  the  reduction  made  by  the  Federal  Reserve  Act  in  the 
amount  of  reserves  to  be  held  against  time  deposits  and  the  further 
reduction  enacted  in  1917.  The  fact  that  the  spectacular  growth  of 
time  deposits  dates  from  that  time  strongly  suggests  a  connection. 
Although  the  circumstance  that  banks  offered  a  higher  rate  of  interest 
on  time  than  on  demand  deposits  should  suffice  to  prove  that  they  thought 
the  former  more  advantageous  than  the  latter,  many  writers  have  argued 
that  the  gain  which  banks  could  expect  from  a  transfer  of  demand  depos- 
its to  time  account  was,  owing  to  the  incident  interest  charge,  small  and 
that  there  was,  hence,  little  if  any  motive  for  them  to  encourage  their 
customers  to  effect  such  transfers.  That  depends  on  what  they  did  with 
the  resulting  increment  in  their  lending  power.  Economists  who  assume 
that  they  would  buy  governments  had  no  difficulty  in  showing  that  the 
gain  was  small.  In  fact,  the  writer  believes  that  in  some  cases  there  must 
have  been  a  trifling  loss.  Smallness  of  gain,  however,  is  frequent  in  the 
banking  business  and  many  a  small  loss  is  undergone  in  order  to  acquire 
or  satisfy  an  otherwise  valuable  customer.  And  if  the  increase  in  lending 
power  went  into  customers'  lines  of  credit  or  into  mortgages — real  estate 
loans  rose  from  less  than  600  millions  in  1919  to  nearly  3.2  billions  in 
1929  (all  member  banks),  a  rate  of  increase  much  greater  than  that  of 
total  earning  assets — there  was  not  only  gain  but  a  very  attractive  one. 
But  this  is  not  the  point.  It  is  quite  sufficient  that  there  was  motive 
on  the  part  of  customers.  About  that  there  cannot  be  any  doubt. 
Moreover,  while  a  nonnegligible  premium  on  time  deposits  was  offered 

variations  in  the  rate  of  spending  and  as  to  show  that  spectacular  growth  of  time  deposits 
could  come  about  without  any  underspending  as  well  as  without  any  saving.  It  follows 
that  it  would  be  dangerous  to  rely  on  time  deposits  as  an  index  of  people's  aversion  to 
spending.  Still,  although  a  demand  deposit  may,  and  often  does,  become  as  idle  as 
any  time  deposit,  it  is  reasonable  to  assume  that  sums  which  are  not  expected  to  be  wanted 
at  all  for  a  time  are  even  more  likely  to  migrate  to  time  account  than  others.  This 
fits  in  very  well  with  the  violent  fluctuations  in  the  rate  of  growth  of  total  time  depos- 
its, which  to  the  writer's  astonishment  Dr.  Currie  flatly  denies  (op.  cit.,  p.  132).  They 
increased  from  December  1921  to  December  1927  (reporting  member  banks;  monthly 
averages  of  weekly  figures)  by  about  $3.4  billions.  Nearly  two-thirds  of  this  increase 
occurred  from  1921  to  1922,  from  1923  to  1924  and  from  1926  to  1927  in  obvious  connection 
with  banks'  investments  (see  Chart  L),  which  in  turn  were  no  less  obviously  related  to 
open-market  purchases  of  the  Federal  system.  These  two-thirds,  at  all  events,  are  pure 
creations  of  monetary  management  and  emerged  on  time  account  because,  in  accordance  with 
our  contention,  emphasized  throughout  this  book,  such  creations  tend  to  become  idle. 
Deposits  of  all  national  savings  banks,  on  the  contrary,  increased  in  fact  much  more 
steadily  (by  about  500  millions  a  year  from  1921  to  1927,  figures  as  of  June  30)  and  are 
"not  affected  to  any  great  extent  by  inflows  and  outflows  of  reserve  funds,"  which  Dr. 
Currie  avers  with  respect  to  time  deposits  (p.  99).  But  for  the  reason  mentioned  in  the 
preceding  note  we  hesitate  to  accept  their  rate  of  increase  as  a  standard  by  which  to  dis- 
tinguish genuine  savings  from  other  time  deposits. 


856  BUSINESS  CYCLES 

to  them  in  the  shape  of  a  rate  of  interest  considerably  higher  than  that 
which  was  paid  on  demand  deposits,  changing  habits  greatly  reduced 
the  incident  inconvenience,  at  least  for  households:  the  greater  the  rela- 
tive importance  in  the  budget  of  expenditure  on  durable  goods,  payment 
for  which  is  lumpy  and  postponable,  and  the  more  widespread  the  prac- 
tice of  charge  accounts,  the  less  need  there  is  for  absolutely  ready  cash  or 
advantage  in  holding  it. 

The  view  that  in  the  twenties  time  and  demand  deposits  were  essen- 
tially the  same  kind  of  thing  is  verified  by  the  relation  to  each  other  of 
the  two  top  curves  on  Chart  XL VI.  From  total  outside  deposits,  banks' 
investments  have  been  deducted  on  the  theory  that  they  tend  to  be 
idle.  In  doing  this  we  are  no  doubt  overshooting  the  mark — for  we  know 
that  even  for  prewar  times  it  would  not  be  true  to  hold  that  banks' 
investments  always  proceed  from  the  banks'  initiative — as  we  are  over- 
shooting it  by  including  all  time  deposits.  All  the  more  interesting  is 
the  result  of  the  experiment.  Our  deposits  curve  fits  the  debits  curve 
excellently  from  1921  to  1927  and  the  different  behavior  of  the  two  curves 
in  1919  and  1920  and  again  in  1928  and  1929  is  easily  accounted  for  by 
the  irregularities  of  those  years,  mainly  by  the  excesses  of  speculation, 
which  would  also  influence  outside  debits.  The  curve  of  outside  demand 
deposits  does  not  fit  so  well  and  thereby  raises  a  typical  spurious  problem, 
viz.,  why  the  "velocity"  of  these  deposits  should  have  so  much  increased1 
and  how  it  was  possible  for  it  to  do  so  or,  to  put  the  matter  differently, 
how  it  was  that  outside  debits  behaved  as  they  did  (increasing  from  17.6 
billions  in  1919  to  27.7  in  1929)  with  demand  deposits  increasing  only 
from  a  little  over  6  billions  to  about  8.2  The  much  more  natural  complex- 
ion of  the  curve  of  "velocity"  of  total  outside  deposits  (see  again  Chart 
XLVI)  will  strengthen  our  belief  that  we  have  in  fact  got  hold  of  the 
approximately  true  figure  and  that  our  diagnosis  of  the  nature  of  time 
deposits  is  correct.3  But  it  must  be  added  that  there  is  another  possible 

1  The  index  of  the  rate  of  turnover  of  demand  deposits  in  principal  cities,  which  has 
been  constructed  by  the  Federal  Reserve  Bank  of  New  York  (1919-1925  =  100),  in  fact 
starts  with  100  at  the  beginning  of  1921  and  after  a  fall  returns  to  it  in  1922,  then  remains 
at  that  level  through  1924,  after  which  it  rises,  with  an  interruption  in  1926,  until  it  reaches 
200  in  1929.     This  includes  New  York.     But  if  we  exclude  New  York,  the  paradox  remains, 
although  toned  down. 

2  Monthly  average  of  deposits  other  than  U.  S.  Government  deposits,  including  Due  to 
minus  (Due  from  plus  Items  in  Process  of  Collection).     Dr.  Currie's  series  of  total  money 
"supply"  (op.  cit.,  p.  83;  including  Cash  outside  Banks;  objections  to  the  term  Supply, 
which  in  this  connection  is  so  pregnant  with  misleading  associations,  need  not  be  repeated) 
displays,  of  course,  the  same  phenomenon.     It  increases  from  nearly  22  billions  in  1921 
(when  total  debits  were  nearly  88.3  billions)  to  nearly  26.7  in  1929  (when  toted  debits  were 
nearly  78  billions).     Since  total  debits  are  heavily  loaded  with  speculative  transactions, 
this  is  not  more  but  less  striking  than  the  comparison  in  the  text. 

8  It  may  be  said  at  once  that  the  same  diagnosis  applies  to  English  deposit  accounts  and 
to  what  corresponds  to  these  in  Germany  although,  owing  to  the  absence  of  the  particular 
incentive  imparted  by  American  banking  legislation,  to  a  lesser  extent. 


1919-1929  857 

reason  why  comparatively  so  small  an  increase  in  outside  demand  deposits 
should  have  proved  adequate.  War  and  postwar  boom  profits  were  not 
annihilated  by  the  subsequent  crash.  Only  an  insignificant  shrinkage  of 
outside  demand  deposits  occurred  in  1921.  There  is  a  difference  of  only 
about  1  billion  between  the  monthly  peak  in  1920  and  the  monthly  trough 
in  1921  and  this  was  quickly  made  up.  In  other  words,  we  see  again  that 
the  monetary  system  of  the  United  States  stayed  expanded  after  the  war 
and  did  not  contract  in  response  to  the  temporary  downward  revision  of 
values.  Thus,  the  latter  left  people  amply  provided  with  owned  funds 
and  for  a  time  at  least,  i.e.,  while  this  slack  was  being  taken  up,  the  eco- 
nomic body  was  to  some  extent  able  to  grow  within  the  existing  coat.1 
However,  taken  at  its  highest  possible  value,  this  explanation  cannot  be 
considered  as  sufficient,  even  to  the  end  of  1924;  while  after  that,  i.e.* 
precisely  for  the  years  when  the  phenomenon  of  increasing  "velocity"  is 
most  clearly  marked,  it  is  not  available  at  all. 

b.  Total  deposits  minus  investment,  then,  unlike  demand  deposits, 
display  during  the  period  the  general  tendency  or  descriptive  trend  that 
prewar  experience  leads  us  to  expect,  though  the  "normality"  of  the 
observed  gradient,  in  this  case  as  it  was  in  that  of  wage  rates,  is  compat- 
ible with  an  abnormally  high  level  owing  to  the  location  of  the  starting 
point.  Accepting  the  latter,  however,  we  also  see  that  fluctuations 
themselves  were  not  less  normal  from  our  standpoint  than  the  descriptive 
trend  they  produced.  In  particular,  we  see  the  footprints  of  the  Juglar 
recovery  and  the  four  phases  of  the  last  Kitchin  of  the  third  Juglar.2  In 
1925  and  after,  the  prosperity  and  recession  of  the  fourth  Juglar  are  well 
marked  by  an  increase  in  our  quantity  and  by  its  tapering  off.  While 
this  does  not  show  in  outside  demand  deposits  alone,  another  feature  of 
special  interest  can  be  observed  just  as  well  in  these:  the  behavior 
of  deposits  with  respect  to  debits,  from  the  last  quarter  of  1929  on. 
While  debits  fell  sharply,  both  outside  demand  deposits  and  total  outside 

1  Some  part  of  the  surplus  funds  which  were  set  free  by  the  reduction  of  operations 
and  the  fall  in  prices  may  have  migrated  to  time  account  very  early  in  the  downturn,  or 
even  before.     This  possibly  explains  part  of  the  strong  increase  in  (total)  time  deposits  in 
1920,  the  monthly  average  of  which  year  is  nearly  800  millions  above  that  of  1919.     The 
general  swerve  toward  the  time  account  must,  however,  not  be  forgotten  and  makes  it 
difficult  to  be  very  positive  on  the  point. 

2  These  we  also  see  in  the  demand  deposits,  which  do  not  otherwise  reflect  well  either 
the  "trend"  or  the  fluctuations.     It  has  been  pointed  out,  in  proof  of  a  fundamental  change 
from  prewar  conditions  and  of  the  dominating  influence  of  open-market  operations  that 
demand  deposits  increased  strongly  in  the  years  of  relapse,  1924  and  1927.     Without 
denying  the  element  of  truth  in  this — even  our  series  is,  of  course,  not  quite  independent  of 
open-market  operations,  although  much  less  affected  by  them — it  should  be  observed 
that  the  deviation  from  prewar  experience  is  so  very  obvious  only  in  yearly  figures  and  total 
demand  deposits.     As  to  the  case  of  1924,  for  instance,  even  total  demand  deposits  reflect 
the  business  situation  quite  faithfully  by  sagging  from  the  middle  of  1928  to  the  middle  of 
1924  in  correspondence  to  loans  and  discounts  (see  below).     And  New  York  demand  depos- 
its are  a  different  matter  and  should  not  be  included. 


858 


BUSINESS  CYCLES 


deposits  minus  investments  fell  much  less.  As  we  saw  at  the  beginning 
of  the  period  that  a  cloud  of  potential  inflation  failed  to  burst,  so  we  see 
now  that  in  1930  business  operations  contracted  in  the  midst  of  a  most 
plentiful  "supply"  of  money — the  coat  refusing  to  shrink  in  proportion 
to  the  shrinking  of  the  body.  The  behavior  of  total  outside  deposits, 
total  deposits,  and  New  York  net  demand  deposits,  and  the  relations  of 
these  scries  to  that  of  outside  debits  should  now  be  studied  (Chart  L). 


NET  DEMAND  DEPOSITS 
NEW  YORK  CITY 


1919     1920     1921      1922     1923     1924     1925    1926     1927    1928    1929     1930    193!      1932    1933     1934 
CHART  L. — United  States  (see  Appendix,  p.  1072). 

The  outstanding  fact  of  very  close  covariation  between  Outside  Loans 
(and  Discounts)  and  Outside  Debits  next  calls  for  our  attention.  Clearly 
the  former  dominated  the  latter  during  our  period  quite  as  much  and  in 
quite  the  same  sense  as  ever.  This  emphatically  normal  state  of  things 
loses  nothing  of  its  importance  by  two  features  which  have  been  much 
stressed  in  the  discussion  of  postwar  banking  figures. 

First,  it  will  be  observed  that  while  dominating  debits,  loans  did 
not  dominate  the  link  between  the  two,  viz.,  deposits.  This  is  obvious 
in  the  case  of  net  demand  deposits,  though  not  so  much  as  some  theorists 


1919-1929  859 

would  have  us  believe,1  but  to  a  lesser  extent  it  is  also  true  of  time  plus 
demand  deposits.  It  is  not  true  only  for  our  quantity,  for  which  covaria- 
tion with  loans  is,  of  course,  tautological.  But  this  circumstance  testi- 
fies precisely  for  our  view  and  against  the  revised  quantity  theories,  for 
it  shows  that,  broadly,  only  those  variations  in  deposits  which  corre- 
sponded to  loans,  turned  into  variations  of  expenditure,  and  that  those 
which  owed  their  existence  to  other  factors  did  not,  whereas  according 
to  those  theories  they  should  have  done  so  just  as  much  as  any  other. 
Once  more,  facts  conspire  to  suggest  that  regulating  deposits  does  not 
imply  regulating  expenditure  or  the  pulse  of  business.  And  this  means 
that  the  new  feature  in  the  behavior  of  deposits  and  the  factors  respon- 
sible for  it  cannot  claim  the  significance  widely  attributed  to  them, 
though,  as  we  shall  see,  they  may  claim  another. 

Second,  it  has  often  been  pointed  out  that,  whatever  total  loans  and 
discounts  may  have  done,  the  commercial  loans  as  represented  by  the 
All-Other  category  of  banking  statistics2  remained  almost  stationary 
throughout  the  period.  It  is  true  that  if  we  pass  by  the  first  years  of  the 
period  and  take  for  a  starting  point  1923,  when  liquidation  of  the  crisis 
may  be  assumed  to  have  been  over,  we  get  an  almost  steady  upward 
movement  and  a  total  increase  amounting  by  1929  to  nearly  2  billions 
for  weekly  reporting  member  banks  in  principal  cities.3  But  if  we 
exclude  real  estate  loans,  the  figure  for  all  member  banks  (June  30) 
never  reaches  the  1921  peak,  and  only  1926,  1928,  and  1929  display  any 
marked  increase.  Even  if  we  do  not  do  this,  the  share  of  all  other  in 
total  loans  declined  from  about  70  per  cent  in  1921  to  below  55  per  cent 
in  1929.  This,  of  course,  is  no  new  tendency  but  only  continues  one  that 
was  present  (in  national  bank  figures)  before  the  war.  It  is  new  only  in 
that  it  now  spread  to  absolute  figures.4  If,  however,  only  quantitative 

1  A.  A.  Young  (op.  cit.,  p.  62)  rightly  emphasizes,  for  national  banks,  "the  general 
correlation  of  the  movements  of  loans  and  discounts  and  of  demand  deposits  both  in  and 
outside  of  New  York  City."     The  divergence,  however,  asserted  itself  mainly  after  the 
date  (or  from  shortly  before  the  date)  at  which  his  discussion  ends.     And  there  was,  of 
course,  considerable  divergence  in  details.     The  more  important  cases  are  associated 
with  the  investment  item.     But  this  explanation  is  not  always  sufficient  and  in  some  cases 
nice  problems  arise  into  which,  however,  it  is  not  possible  to  enter  here. 

2  Let  us  recall  that  the  best  that  can  be  said  for  that  category  is  that  it  has  more  to  do 
with  what  is  usually  referred  to  as  commercial  credit  than  others.     But  many  loans  on 
securities  are  no  less  genuinely  business  loans.     The  greater  part  of  loans  on  real  estate 
is  not. 

3  See  the  Monthly  Review  of  the  Federal  Reserve  Bank  of  New  York  for  Oct.  1,  1932,  p. 
78.     New  York  is  included.     Professor  Mills  (op  cit.,  p.  450)  gives,  for  all  reporting  member 
banks,  also  including  New  York,  an  average  annual  rate  of  increase  of  8.2  per  cent,  for 
loans  on  securities  10  5  per  cent,  between  1922  and  1929. 

4  Dr.  Anderson  (Chase  Economic  Bulletin  for  Apr.  8,  1927,  p.  18)  drew  attention  to  the 
low  percentage  of  total  member  bank  assets  formed  by  paper  eligible  in  the  technical  sense. 
Dr.  Currie  also  stresses  this  point  (op.  cit.,  p.  117).     The  theoretical  schema  underlying  the 


860  BUSINESS  CYCLES 

and  not  qualitative,  the  change  is,  within  the  secondary  order  of  impor- 
tance that  attaches  to  the  figures  and  processes  of  the  sphere  of  credit,  a 
fundamental  one  and,  hence,  requires  careful  consideration. 

One  thing  is  clear:  it  did  not  originate  with  banks  since  these  stand  to 
lose  by  it  what,  in  the  long  run,  is  their  most  profitable  line  of  business. 
It  must  have  been  the  customers  who  turned  away  from  this  method  of 
financing.  Among  them  we  can  discard  the  households,  for  these  con- 
tinued to  borrow  freely — in  fact,  more  freely  than  ever — as  the  develop- 
ment of  real  estate  loans  and  of  personal  loan  departments  suflacesto  show. 
They  also  borrowed  indirectly  from  banks,  through  various  types  of 
financing  agencies  which  discounted  with  banks,  and  through  helping 
to  create  installment  paper,  which  also  formed  an  increasing  percentage 
of  all  other  loans.  Nor  can  the  small  and  medium-sized  trading  and 
manufacturing  business  be  responsible  for  the  change,  for  this  remained, 
much  as  it  had  always  been,  dependent  upon  the  customers'  line  of  credit 
for  all  its  needs,  the  financing  of  essentially  long-term  investments 
included.  Corporate  business,  hence,  is  what,  in  accordance  with 
common  knowledge  and  general  opinion,  emerges  from  this  process  of 
elimination.  But  why  should  bigger  and  big  business  have  taken  this 
new  departure?  Simply  because  the  plethora  of  money  made  it  easy 
and  profitable  to  embark  upon  a  course  which  in  itself  appeals  to  execu- 
tives who  are  always  jealous  of  anything  that  involves  a  certain  amount 
of  supervision  and  who  for  this  reason  never  love  their  banking  connec- 
tions. As  stated  above,  in  the  first  years  of  the  period  the  more  successful 
concerns  had  at  their  disposal  ample  funds  with  which  to  finance  them- 
selves. These  had  been  previously  assembled  from  profits1  and  in  many 
cases  preserved  by  timely  withdrawal  from  the  firing  line.  Later  on, 
when  bond  yields  declined  and  stock  markets  boomed,  money  flowed  so 
easily  toward  corporate  industry,  at  rates  with  which  no  bank  could 
compete  that,  taking  full  advantage  of  this  situation,  concerns  became, 
in  the  process  of  expansion,  creditors  rather  than  debtors  as  far  as  ready 
money  is  concerned,  keeping  financially  ahead  of  requirements  and 
eventually  entering  the  great  depression  with  a  financial  outfit  which  was 
nothing  short  of  luxurious.2  Some  of  them  were  even  able  to  finance 

Federal  Reserve  Act  increasingly  failed  to  fit  American  conditions.  The  former  author 
dislikes,  the  latter  likes  this  development,  but  for  us  it  is  only  the  facts  that  matter. 

1  We  may  recall  that  undivided  profits  of  corporations,   unconnected  for  difference 
between  depreciation  at  cost  and  at  current  prices,  amounted  to  4,310  millions  in  1919,  a 
figure  never  reached  again.     But  financing  from  profits  of  course  continued. 

2  The  cash  item  of  corporate  balance  sheets  reflects  this  imperfectly,  but  it  does  reflect  it. 
Its  percentage  relation  to  total  assets  can  be  calculated  from  the  corporation  tax  material, 
from  1926  on.     Including  with  cash  tax-exempt  securities,  which  served  as  temporary 
investments,  and  deducting  from  this  item  for  All  Corporations  the  cash  plus  tax-exempt 
investments  held  by  banks  and  financial  corporations,  we  get  for  industrial  and  commercial 


1919-1929  861 

their  investments  in  1935  and  1936  by  funds  that  had  been  raised  during 
the  speculative  mania  of  1928  and  1929, 1 

Now,  that  "  money  '*  still  came  from  or  come  through  banks.  It  was 
to  a  great  extent  created  by  them,  exactly  as  it  would  have  been  had 
corporations  directly  borrowed  from  them.  Only  it  was  not  created 
through  loans  to  industry,  which  would  have  swelled  other  loans,  but 
through  loans  to  buyers  of  bonds  and  stocks,  which  swelled  loans  on 
securities.  At  least  this  is  that  alternative  to  lending  directly  to  indus- 
try which  will  bring  out  most  clearly  the  point  we  are  trying  to  make. 
Banks,  of  course,  also  bought  bonds — and  stocks  through  their  affiliates — 
and  placing  these  acquisitions  became  so  important  that  a  new  type  of 
bank  executive  emerged  who  had  little  of  the  banker  and  looked  much 
like  a  bond  salesman.  But  this  we  will  put  aside  for  the  moment  in  order 
to  focus  attention  on  the  fact  that  to  a  considerable  extent  credit  creation 
by  loans  on  securities  was  a  substitute  for  "business  loans" — hence,  to 
the  same  extent  not  a  net  addition  to  total  volume  of  balances — and, 
inasmuch  as  it  was,  served  the  same  purpose  and  not  an  additional  one. 
It  does  not  follow  that  the  actual  amount  of  balances  created  was  the 
same  as  that  which  would  have  been  created  by  the  orthodox  method 
or  that  the  change,  being  one  of  mere  technique,  was  unimportant.  On 
the  contrary,  it  is  safe  to  say  that  in  the  midst  of  rioting  stock  markets 
creation  went  on  at  a  pace  very  different  from  that  which  would  have 
been  set  in  bankers'  conference  rooms.  The  steering  and  balancing  parts 
of  the  capitalist  machine  were  seriously  and  perhaps  permanently 
impaired.  From  remedies  eventually  applied  to  the  ensuing  situation  a 
novel  apparatus  of  banking  may  well  evolve  from  which  there  is  no  road 
back  to  what,  from  the  standpoint  of  our  model,  would  have  to  be  called 
normal  conditions.  Nonetheless  it  is  necessary  to  take  account  of  that 
substitution  of  one  method  of  deposit  manufacture  for  another  if  we  wish 
to  speak — the  writer  does  not  care  whether  we  do  or  not — of  the  "infla- 
tion" of  the  twenties.  And  for  the  immediate  purpose  in  hand  it  is 

corporations  the  following  amounts  in  billions  (we  add  the  years  of  the  great  depression) : 
1926,  9.9;  1927,  10.3;  1928,  11.1;  1929,  10.9;  1930,  10.4;  1931,  9.1;  1932,  9.1;  1933,  8.8;  or 
in  percentages  of  total  assets:  5.65,  5.86,  6.13,  5.57,  5.37,  5.20,  5.37,  5.41  per  cent — a  series 
the  remarkable  stability  of  which  is  well  worth  noticing. 

1  Acceptances  and  commercial  paper  bought  by  banks  are,  to  September  1934,  included 
in  Other  Loans.  But  from  the  standpoint  of  the  relations  between  corporate  industry 
and  banks,  these  items  should  also  be  excluded  and  listed  among  the  methods  of  inde- 
pendent financing.  For  big  corporations  either  directly  or — through  intermediate  agencies 
owned  by  them,  such  as  acceptance  corporations — indirectly  appealed  to  the  open  market 
much  as  the  Federal  Treasury  did.  And  although  banks  were  faute  de  mieux  forced  to 
buy  such  paper,  these  purchases  do  not  constitute  "  discounting  for  customers."  This 
practice,  in  fact,  meant,  in  the  short-term  field,  as  complete  a  rupture  of  the  old  bank- 
customer  relation  as  bond  issues  meant  in  the  long-loan  field. 


862  BUSINESS  CYCLES 

important  to  recognize  the  same  old  process  of  financing  enterprise  in 
the  new  garb,  the  material  of  which  includes  the  security  loans,  and  to 
see  both  the  stagnation  of  "commercial"  and  the  expansion  of  security 
loans  in  this  light. 

Total  outside  loans  and  discounts,  then,  are  the  really  relevant  figure. 
Their  share  in  total  earning  assets  varied  from  1922  to  1928  between  71 
and  74  per  cent.1  And  there  is  nothing  obviously  abnormal  or  unprece- 
dented in  their  average  rate  of  increase — their  "descriptive  trend" — 
during  the  period.  They  also  move  in  the  Kitchins,  and  show  the  rise 
of  the  fourth  Juglar,  according  to  expectation.  Divergences  in  the 
behavior  of  the  components  give  rise  to  various  problems  of  detail  into 
which  we  cannot  enter.  First  differences  of  the  original  items  of  the 
all-other  loan  series  reflect  short-run  business  situations  more  faithfully 
than  first  differences  of  the  original  items  of  the  security  loan  series — 
which  is  easy  to  understand  on  grounds  of  financial  technique — so  that 
the  former  series  remains,  for  some  purposes,  more  useful  than  the  latter. 

c.  Turning  now  to  members'  investments,2  we  will  avail  ourselves 
of  the  opportunity  to  look  at  some  of  the  developments  discussed  under 
(a)  and  (b)  from  another  standpoint,  viz.,  the  standpoint  of  the  banks. 
Parallelism  between  loans  and  deposits  has,  of  course,  never  been  perfect 
but  always  interfered  with  by,  first,  the  efflux  from  and  the  influx  into 
banks  of  currency;  second,  the  efflux  from  and  the  influx  into  the  country 
of  the  monetary  metal  (including  efflux  into  and  influx  from  the  arts) ; 
and,  third,  banks'  investments.  During  the  period  under  discussion, 
none  of  these  factors  operated  as  it  did  before  the  war.  But  the  change 
that  primarily  impressed  theorists  and,  in  fact,  amounted  to  a  revolution 
in  banking  practice  and  in  the  structure  of  earning  assets  occurred  in  the 
absolute  and  relative  importance  of  the  third  (see  Chart  L). 

This  change  dates  far  back.  As  we  know,  it  is  clearly  indicated  in 
national  bank  figures  ever  since  the  nineties,  and  that  theoretical  inter- 
pretation to  which  we  have  grown  accustomed — the  investment  theory 
of  banking — could  have  been  inspired  nearly  as  well  by  the  course  of 
events  from  1899  to  1908,  when  "secured"  loans  and  investments  behaved 

1 1920,  1921,  and  1929  display  higher  pecentages.  It  is  of  some  interest  to  note  that 
there  is  a  rough  covariation  between  the  year-to-year  changes  in  these  percentages  and  the 
"velocity"  of  deposits;  between  the  ratio  of  loans  and  discounts  to  net  demand  deposits 
and  the  commercial  paper  rate;  and  between  the  latter  and  the  ratio  of  loans  to  banks' 
investments.  Interpretation  from  our  standpoint  is  too  obvious  to  detain  us. 

2  The  figures  plotted  on  Chart  L  are,  in  fact,  those  of  (weekly  reporting)  member  banks 
outside  New  York  (see  Appendix).  But  the  reader  should  remember  that  within  our 
general  reason' ng  the  term  Member  Bank  or  simply  Member  has  a  technical  meaning  of 
our  own.  Weekly  reporting  member  banks  outside  of  New  York  are,  on  the  one  hand, 
only  a  sample  of  a  more  comprehensive  universe  designated  by  this  term  in  its  usual 
meaning  and  for  us,  on  the  other  hand,  only  the  real-world  representatives  of  a  theoretical 
entity  (=  all  noncentral  banks). 


1919-1929  863 

in  a  manner  highly  suggestive  of  postwar  developments.  The  Federal 
Reserve  Act  and  the  amendment  of  1917,  on  the  one  hand,  and  the  gold 
influx,  on  the  other,  providing  additional  facilities,  war  finance  imparted 
an  impulse1  which  would  not  in  itself  have  meant  more  than  a  very 
natural  temporary  deviation,  under  the  pressure  of  exceptional  circum- 
stances, of  no  permanent  importance  for  the  financial  mechanism  of 
cycles.  We  observe,  in  fact,  that  as  soon  as  those  circumstances  had 
passed,  banks  were  anxious  to  normalize  their  position  and  to  reduce  their 
share  in  the  iceberg  that  was  swimming  about  and  encumbering  naviga- 
tion in  the  money  market.  Since  this  was  done  during  1919  and  1920, 2 
when  the  postwar  boom  was  in  full  swing,  we  may  infer  that,  while 
maneuvering  back  to  normal  conditions,  they  were  also  maneuvering 
for  room  to  satisfy  industrial  and  commercial  requirements.3  But  this 
endeavor  to  eliminate  what,  at  the  beginning  of  the  period,  they  (and 
also  the  Federal  Reserve  Board)  seem  to  have  felt  to  be  an  abnormal  and 
undesirable  situation,  was  soon  given  up  and  never  resumed.  By  the 
end  of  1922  total  investments  were  almost  back  to  the  peak  figure,4  to 
increase  to  ever  new  levels  for  the  rest  of  the  period.6 

Interpretation  must  start  by  observing  that  within  this  descriptive 
trend  fluctuations  correspond,  at  least  as  to  direction  and  timing,  to 
expectation  from  our  model.  By  increasing  their  investments  in  the 
depression  of  1921  and  in  1922,  by  decreasing  them  in  the  course  of  1923, 
increasing  them  again  in  1924,  keeping  them  substantially  steady  during 
1925  and  1926,  increasing  them  in  1927,  and  mildly  decreasing  them  in 
1928  and  1929,  banks  substantially  conformed  to  the  prewar  pattern  of 
behavior:  there  was  fairly  satisfactory  inverse  covariation  between  rates 
of  change  in  investments  and  in  loans  which,  particularly  visible  after 

1  "All  Banks"  increased  their  investments  from  about  5.5  billions  in  1914  to  about  11.9 
in  1919   (see   15th  annual  report  of  the  Federal  Reserve  Board,  p.  111).     The  national 
banks  up  to  1917  confined  their  holdings  of  governments  substantially  to  what  was  neces- 
sary to  cover  their  note  circulation  and  government  deposits.     Those  in  New  York  City 
took  the  great  stride  in  that  year,  those  outside  New  York  took  it  in  1917  and  1918  and 
reached  the  maximum  at  the  beginning  of  1919. 

2  New  York  banks  also  liquidated  other  than  government  securities.     Outside  banks 
did  so  only  to  a  quite  insignificant  extent  and  only  during  a  few  months. 

3  And  in  fact  a  very  substantial  increase  in  loans  occurred,  as  has  been  noticed  in  Sec. 
E.     As  has  been  stated   there  and  will  be  repeated  again — it  can  never  be  repeated 
enough — this  is  very  material  to  any  rational  appraisal  of  that  "deflation."     In  1921  the 
position  of  banks  was  further  strengthened  by  a  decrease  of  currency  in  circulation  which 
amounted  to  940  millions,  by  an  increase  in  treasury  currency  of  nearly  220  millions  and  by 
the  gold  influx  (749  millions). 

4  Outside  banks  had,  however,  only  reached  about  two-thirds  of  the  peak  figure  of 
their  holdings  of  governments  by  the  end  of  1922. 

6  The  figures  of  June  30,  1928,  after  which  they  declined  for  about  2  years,  were  17.8 
billions  for  All  Banks.  Investments  of  All  Member  Banks  were  about  6  billions  in  1921 
(June  80)  and  nearly  10.76  billions  in  1928;  10.05  in  1929. 


864  BUSINESS  CYCLES 

the  elimination  of  that  descriptive  trend,  indicates  that  the  former  must 
have  retained  something  of  their  old  role.  We  will  add  the  complemen- 
tary fact  that  members  (in  our  sense)  also  conformed  to  the  prewar 
pattern  in  that  they  continued  to  look  primarily  after  their  customers' 
credit  and  to  make  it,  even  at  temporary  sacrifices,  the  pivot  of  their 
business.1  No  grasp  of  banking  developments  in  the  twenties  is  possible 
unless  this  be  clearly  realized. 

But  that  descriptive  trend  in  investments  is  entirely  due  to  the  three 
spurts  in  1922,  1924,  and  1927,  which  are  obviously  associated  with 
the  three  major  buying  campaigns  of  the  reserve  system.2  It  should 
be  noticed,  however,  that  this  relation  is  complicated  by  a  number  of 
other  factors,  and  also  that  only  in  the  second  case  did  the  action  of  the 
Federal  Reserve  Bank  of  New  York  precede  the  purchasing  operations  of 
member  banks.  In  the  two  other  cases  the  latter  moved  first,  which 
should  be  sufficient  to  convince  us  that  these  spurts  contained  an  element 
which  was  independent  of  Federal  reserve  action:  this  action  supplied 
an  additional  impetus  and  additional  liquid  means,  but  was  not  the  sole 
causa  efficiens.  This  is  also  corroborated  by  the  fact  that  the  net  result 
of  the  reserve  system's  major  dealings  in  governments,  from  January 
1922  to  October  1929,  was  minus  65  millions.3  Nevertheless,  it  was  one 
important  factor4 — the  second  or,  if  we  count  the  war  effect,  the  third — 
in  shaping  members'  investments. 

The  third  or  fourth  factor  is  the  most  interesting  one.  It  consists 
of  the  conditions  of  the  banks'  customer  business  adverted  to  above, 
III,  b.  When  banks  in  and  after  1922  saw  this  business  floating  away 
from  them  on  the  tide  of  abundant  money,  they  did  not,  strictly  speaking, 
acquire  a  new  motive  to  look  to  investments  for  a  permanent — i.e.,  more 

1  Member  banks  (in  the  official  sense)  are,  since  October  1928,  required  to  give  informa- 
tion which  enables  us  to  distinguish  customers'  from  open-market  loans.     This  information, 
published  in  the  Bulletin  and  the  Annual  Reports  of  the  Federal  Reserve  Board  is  very 
revealing  and  has  been  drawn  upon  in  what  follows  in  our  text.     The  point  which  is 
material  at  the  moment  is  that  banks  did  not  turn  to  temporary  investment  in  the  open 
market  at  the  expense  of  their  loans  to  customers  when  this  was  more  remunerative,  as  it 
was  in  1929.     This  implies  that  selling  and  purchasing  bonds  must  also  have  been,  for  them, 
secondary  or  subsidiary  to  customers'  loans. 

2  But  cp.  the  discussion  of  the  modus  operandi  of  open-market  operations  which  follows 
below,  sub  V. 

3  There  is  always  a  danger  of  overestimating  the  causal  role  of  Federal  reserve  purchases, 
since  the  system  would  naturally  buy  in  a  situation  in  which  member  banks,  finding  their 
funds  idling  in  their  hands,  would  increase  their  investments  independently  of  any  action 
by  the  system.     This  will  presently  become  still  clearer. 

4  The  present  writer  entertains  a  profound  respect  for  any  opinion  held  by  that  great 
economist,  A.  A.  Young,  and  records  dissent  with  reluctance,  but  he  cannot  understand 
how  Young  can  have  held  (op.  cit.,  p.  59)  that  the  open-market  purchases  of  the  Federal 
Reserve  Bank  of  New  York  "appear  to  have  no  distinguishable  effects"  on  the  amount  of 
government  securities  held  by  New  York  national  banks. 


1919-1929  865 

than  temporary — source  of  employment  for  their  deposit  manufacturing 
facilities,  for  as  we  have  seen,  this  situation  was  not  absent  from  the 
prewar  picture.  But  the  old  motive  gathered  additional  weight.  And 
so  they  increasingly  took  to  holding  and  handling  bonds,  outside  banks 
in  particular  other  than  government  bonds,  which  they  acquired  steadily 
even  in  1923.  The  holding  provided  a  return  which,  though  not  very 
attractive  and  falling,  was  supplemented  by  capital  gains  and  gains 
from  the  handling,  especially  the  placing.  So,  while  still  regulating 
their  course  in  the  way  described  above,  they  systematically  bought  more 
when  buying  than  they  sold  when  selling  and  this  is  the  fundamental 
explanation  of  the  descriptive  trend  we  observe.  But  it  also  bears  on 
fluctuations  and  on  the  relation  of  this  series  to  others,  for  inasmuch  as 
the  method  of  financing  industrial  and  commercial  requirements  by  the 
issue  of  new  or  by  the  sale  of  old  bonds  was  increasingly  resorted  to  on 
the  initiative  of  industrial  and  commercial  concerns,  i.e.,  inasmuch  as 
it  increasingly  took  the  place  of  direct  borrowing  from  banks,  a  com- 
ponent was  inserted  that  would  be  positively  and  not  negatively  associated 
with  cyclical  phases.  We  have  observed  this  phenomenon  in  our  survey 
of  prewar  times  but  it  became  much  more  important  during  the  twenties. 
We  shall  expect  it  to  assert  itself  especially  in  the  holdings  of  other  than 
government  bonds.  So  it  does — for  outside  banks  it  seems  to  correlate 
with  business,  for  New  York  banks,  with  speculative  activity,1  as  it 
should.  And  this  explains  much  of  what  is  unsatisfactory  in  the  negative 
association  referred  to  above. 

But  this  negative  association  was  only  weakened  and  not  destroyed 
by  that  component.  So  far  as  it  was  not,  we  shall  expect  some  covaria- 
tion between  investments  and  time  deposits,  since  investments  due  to  the 
initiative  of  banks  are  likely  to  create  idle  deposits  and  idle  deposits  are — 
in  spite  of  the  qualifications  following  from  our  discussion  above — more 
likely  to  be  placed  on  time  account  than  are  others.  Time  deposits 
increased  at  a  much  greater  rate,  which  accords  with  our  theory  that 
they  do  not  primarily  consist  of  idle  funds,  but  represent  a  new  way  of 
handling  certain  classes  of  circulating  funds,  as  does  the  fact  that  the 
logarithmic  ratio  between  investments  and  time  deposits  almost  steadily 
declined.  But  still  there  is  more  than  a  suggestion  of  covariation,  as 
especially  in  New  York  figures  for  1921-1922  and  for  1924.  This  is  what 
remains  of  the  grossly  erroneous  theory  that  banks  invest  the  time 
deposits  with  which  they  are  "entrusted"  by  savers.  We  know  that 
1  It  has  been  observed  by  A.  A.  Young  (op.  cit.,  p.  58)  that  the  security  holdings  of 
New  York  national  banks  display  a  variability  and  quickness  of  response  greater  than 
those  of  outside  national  banks.  "  The  way  in  which  that  [the  New  York]  market  serves 
as  a  distributing  center,  holding  securities  temporarily  until  they  can  be  absorbed  else- 
where, is  reflected  in  the  character  of  some  of  the  fluctuations/'  This  also  applies  to  1927 
to  1929. 


866  BUSINESS  CYCLES 

there  is  at  best  fractional  truth  in  the  implied  relation  between  time 
deposits  and  savings  and  that  a  bank  as  a  matter  of  principle1  can  no 
more  be  said  to  lend  or  invest  a  time  than  a  demand  deposit,  though  to 
the  banker,  provided  he  adheres  to  the  old  phraseology  of  his  trade,2  it 
may  well  look  as  if  he  did.  Moreover,  governments  were  among  the 
most  liquid  of  assets  and  the  argument  that  it  would  be  bad  banking  to 
buy  them — or  any  bonds — except  so  far  as  they  are  balanced  by  slow 
liabilities,  sounds  strange  indeed  when  it  comes  from  bankers  who  felt 
no  compunction  when  entangling  themselves  in  mortgages. 

A  number  of  additional  points  would  have  to  be  dealt  with  in  order 
to  complete  the  picture.  We  confine  ourselves  to  a  few  of  the  most 
important  ones.  First,  banks  improved  their  balance  sheets  by  increas- 
ing their  capital.  This  conversion  of  deposits  into  bank  stock  must 
always  be  kept  in  mind  in  any  study  of  the  behavior  of  deposits  and  in 
discussions  about  the  presence  or  absence  of  "inflation."  It  will,  how- 
ever, suffice  to  notice  the  increase  which  occurred  between  1917  and 
19213  and  which  parallels  other  efforts  to  normalize  their  status  in 
1919  to  1921,  in  particular  the  partial  liquidation  of  investments.  Sec- 
ond, the  Federal  Reserve  Act  was  bound  to  change  the  relation  between 
pure  member  banks  (our  sense)  and  member  banks  that  also  filled 
central  bank  functions  and,  as  far  as  the  former  are  roughly  represented 
by  outside  and  the  latter  by  New  York  banks,  to  affect  the  relations 
between  these  two  groups.  Nevertheless,  these  relations  persisted 
rather  more  than  we  might  have  expected:  the  New  York  correspondent 
retained  to  a  considerable  extent  his  role  in  the  operations  of  the  banks 
outside.4  But  the  transfer  of  reserve  deposits  to  the  Federal  reserve 
banks  affected  interbank  deposits.  The  prewar  relation  between  deposits 
in  New  York  banks  and  money  in  all  banks  is  no  longer  observable. 
Something  is  left  of  the  prewar  relations  between  New  York  and  outside 
net  demand  deposits  and  between  New  York  and  outside  loans  and 

1  The  case  of  loans  on  account  of  others  will  be  mentioned  later  on. 

2  The  same  excuse  cannot  be  pleaded  for  theorists  who  in  all  other  respects  have  ab- 
sorbed the  theory  of  "credit  creation."     To  retain  that  vestige  of  old  doctrine  is  in  fact 
quite  illogical. 

3  Capital  and  surplus  of  national  banks  stood  at  1,845  millions  in  1917  and  that  of 
other  banks,  excluding  savings  banks,  at  1,953  millions.     In  1921  capital  and  surplus  of 
national  banks  had  risen  to  2,300  millions,  capital  and  surplus  of  other  banks  to  2,720 
millions.     But  increase  continued  and  in  1928-1929  helped  to  enhance  the  banks'  lending 
power:  capital,  surplus,  and  undivided  profits  of  member  banks  increased  by  about  2 
billions  from  1922  to  1929,  which  item  plus  total  deposits,  of  course,  about  equals  the 
increase  in  loans  plus  investments  plus  reserves. 

4  This  is  shown  by  the  fact,  among  other  things,  that  rediscounting  with  correspondents 
(New  York  and  others)  went  on,  though  on  a  modest  scale,  throughout  the  period.      See  on 
this  and  cognate  points  S.  E.  Harris,  Twenty  Years  of  Federal  Reserve  Policy,  vol.  II, 
App.  C,  especially  the  chart  on  p.  773. 


1919-1929  867 

discounts.  Other  points  have  been  noticed  already.  Third,  though  the 
new  organ  of  central  banking  and  its  policy  will  be  discussed  later,  its 
influence  on  the  behavior  of  banks  must  be  glanced  at  now. 

It  was  to  be  expected  that  the  presence  of  so  accessible  a  source  of 
additional  "funds"  would  have  saved  member  banks  much  of  their 
trouble  about  liquidity  and,  hence,  profoundly  changed  their  policy. 
In  the  argument  of  mechanistic  theories  about  the  regulation  of  the 
"supply"  of  money,  the  assumption  that  it  did  forms,  in  fact,  an  essential 
link.  And  the  growth  of  real  estate  loans  and  other  symptoms  actually 
point  in  that  direction.1  But,  no  matter  whether  by  compulsion  from 
the  federal  reserve  authorities,  by  tradition,  or  by  perception  of  the 
long-run  advantages  of  the  observation  of  "sound"  principles,  the  banks 
yielded  to  such  temptation  as  there  was  much  less  than  one  might 
think.  To  be  sure,  their  behavior  varied  greatly,  especially  as  between 
first-  and  second-rate  institutions.2  The  recurrence  in  the  Federal 
Reserve  Board's  annual  reports  of  passages  suggestive  of  difficulties 
with  banks  which  tended  to  lean  on  the  snaffle  in  order  to  increase  their 
working  resources  or  at  least  to  profit  from  differences  between  open- 
market  and  rediscount  rates,  is  no  doubt  significant.  On  the  whole, 
however,  there  cannot  have  been  very  much  of  this.  The  modal  and 
still  more  the  high-class  bank  did  not  relish  being  indebted  to  its  reserve 
correspondent  and  was  usually  anxious  to  reduce  its  debt.3  Looking, 
as  has  been  remarked  above,  primarily  after  its  customers'  business,  it 
more  majorum  so  regulated  open-market  commitments  and  investments 
as  to  be  able  to  finance  it  without  falling  back  upon  rediscounting  save  in 
exceptional  circumstances.  This  can  be  seen  clearly  only  since  October 
1928,  when  the  necessary  information  first  becomes  available:  in  the 
last  year  of  the  period  member  banks  in  the  aggregate  actually  increased 
their  customers'  loans,  while  at  the  same  time  paying  back  that  incre- 
ment of  debt  which  they  had  incurred  at  the  reserve  banks  in  the  first 
half  of  1928,  and  they  achieved  this,  until  October  1929,  in  the  face  of  the 

1  In  this  connection  it  should  be  mentioned,  however,  that  member  banks'  reserve 
balances  increased,  January  1922  to  January  1927,  by  nearly  560  millions  (though  "  Federal 
reserve  credit  outstanding"  declined  by  140  and  currency  in  circulation  increased  by  375 
millions)  largely  as  a  result  of  gold  imports.     Reserve  balances  were   (monthly  average) 
1.742  billions  in  1919,  1.655  billions  in  1921,  and  2.375  billions  in  1929.     Some  authors 
regard  this  increase  of  about  700  millions  as  the  pivotal  figure  in  the  monetary  processes 
of  the  period,  in  particular  as  proof  positive  of  the  presence  of  "inflation." 

2  Since  the  largest  banks  happened  to  be  located  in  New  York,  that  difference  may, 
roughly  and  in  the  aggregate,  be  said  to  correspond  to  a  difference  in  behavior  of  New  York 
and  outside  banks.     See  L.  Currie,  op.  cit.>  pp.  91  et  seq. 

3  In  the  case  of  New  York  banks,  increase  in  indebtedness  is,  in  fact,  regularly  followed 
by  a  decrease  in  net  demand  deposits,  as  Dr.  Currie  has  pointed  out  (op.  cit.,  p.  93).     But 
it  must  not  be  forgotten  that  there  are  also  other  reasons,  originating  in  the  cyclical  rhythm, 
which  would  tend  to  produce  that  result,  irrespective  of  reluctance  to  being  in  debt. 


868  BUSINESS  CYCLES 

reserve  banks'  sales  of  governments,  by  reducing  their  loans  to  brokers 
and  their  holdings  of  acceptances  and  commercial  paper,  though  the 
simultaneous  influx  of  gold  greatly  helped.  But  the  behavior  of  redis- 
counts (see  below)  suggests  that  the  inference  may  be  generalized  so  as 
to  cover  the  whole  period.1  It  follows  that  banks,  assisted  no  doubt  by 
exceptionally  favorable  circumstances,  were  never  "loaned  up"  except 
in  the  irrelevant  sense  of  that  phrase  which  includes  secondary  reserves. 
In  this  irrelevant  sense  they  may  have  been  much  of  the  time,  though 
even  in  the  first  quarter  of  1929  there  were  small  excess  reserve's.  They 
never  were  loaned  up  in  the  relevant  sense,  i.e.,  in  the  sense  that  without 
open-market  operations  they  would  have  been  embarrassed  by  a  customer 
asking  for  a  loan.  We  therefore  conclude  that  member-bank  practice 
and  hence  the  meaning  of  member-bank  figures  after  all  differed  less 
from  prewar  patterns,  and  that  the  figures  are  more  amenable  to  interpre- 
tation in  terms  of  the  processes  embodied  in  our  model  than  we  might 
have  expected.  This  must  be  borne  in  mind  when  we  try  to  appraise 
the  role  in  the  processes  of  the  twenties  of  the  increase  of  nearly  700  mil- 
lions (1922  to  1929)  in  the  reserves  of  all  member  banks,  or  of  the  increase 
of  about  11.5  billions  in  their  total  loans  plus  investments. 

d.  Discussion  of  the  struggle  of  German  banks  with  peculiarly 
German  postwar  conditions  would  present  many  points  of  interest  but 
would  not  add  materially  to  our  understanding  of  the  postwar  cyclical 
process.  The  most  important  feature,  foreign  credits,  has  been  fully 
dealt  with  earlier  in  this  chapter.  These,  of  course,  powerfully  influenced 
the  behavior  of  "deposits  and  circulation"  (Chart  XL),  which  also 
reflects  gradual  return,  after  inflation,  to  normal  habits  of  holding  bal- 
ances and  the  Juglar  phases.  "Creditors"  in  leading  banks  (only  very 
roughly  comparable  to  demand  plus  time  deposits  in  the  United  States), 
which  had  been  as  low  as  3.8  billion  reichsmarks  in  1924,  rose  to  about 
12  billions  in  1929  but  there  is,  under  the  circumstances,  nothing  remark- 
able in  this  increase  and,  in  particular,  no  indication  about  the  rate  of 
saving;  if  anything  it  indicates  disproportionate  expansion  of  monetary 
purchasing  power.  The  Debitoren  in  current  account,  with  which  we 
include  advances  and  bills  held  (this  is  very  roughly  equivalent  to  all 
other  loans),  increased  steadily,  though  with  the  expected  variations  in 
rate,  from  a  little  over  3.3  billions  in  1924  to  nearly  9.4  in  1929  and  show 
that  orthodox  bank  credit  did  not  in  Germany  lose  nearly  so  much 
ground  as  it  did  in  this  country.  The  same  can  be  inferred  from  the 

1  The  above  is  perfectly  compatible  with  some  covariation  between  rediscounts  and 
outside  loans  and  discounts  which,  but  for  open-market  operations,  would  be  still  more 
clearly  marked.  They  rose  together  in  the  postwar  boom,  fell  together  to  the  middle  of 
1922,  rose  together  in  the  revival  of  1922-1923  and  then  again  in  1925  and  1928.  Only 
1924  shows  marked,  1927  a  less  marked,  divergence  that  is  easily  accounted  for. 


1919-1929 


869 


series,  which  continued  to  be  of  fundamental  importance,  of  Bills  Drawn, 
although  their  share  in  the  total  of  credit  outstanding  rapidly  declined.1 
Money  in  circulation  also  displays  a  rising  tendency  to  1929  but  surpassed 
the  1913  figure  only  insignificantly  in  that  year.  Other  aspects  will  be 
noticed  later 


1919     1320     1921      1922     1923     1924    1925     1926    1927    1928     1929    I9SO     1931      1932     1933    1934     1935     1936 
CHART  LI. — London  clearing  banks'  figures  (see  Appendix,  p.  1073). 

All  that  it  is  essential  for  our  purpose  to  observe  about  the  English 
case  can  be  seen  at  a  glance  in  Chart  LI. 

The  striking  difference  between  this  picture  and  the  American  and 
German  ones,  due  to  the  "  classic  "  lines  of  the  former,  needs  no  comment 
beyond  a  reminder  that  those  lines  were  the  result  not  only  of  the  mone- 
tary and  fiscal  policy  pursued  but  also  of  properties  of  the  English 
economic  process  which  were  completely  independent  of  them.  Treasury 

1  The  amount  of  bills  held  by  banks  increased  almost  steadily  whether  the  total  in- 
creased or  decreased.  This  shows,  if  proof  be  wanted,  how  very  liquid  they  were  all  the 
time. 


870  BUSINESS  CYCLES 

Bills  Discounted  have  been  inserted  in  order  to  avoid  missing  the  fact 
that  public  financing  was  throughout  the  central  factor  in  the  money 
market  around  which  everything  else  turned.  Advances  display  a  most 
orthodox  inverse  covariation  with  investments,  which  during  the  period 
never  acquired  the  significance  of  their  American  counterpart.  Be  it 
repeated  that  current  accounts1  kept  up  in  1920  and  the  first  half  of  1921, 
while  a  clearings  figure  which  is  supposed  to  parallel  American  outside 
debits  declined  to  a  new  and  very  stable  level.  Moreover,  they  display 
but  the  most  gentle  suggestion  of  cyclical  fluctuations:  Juglar  phases  are 
just  recognizable  in  the  fact  that  they  slowly  fell  to  1925,  then  ceased  to 
fall,  and  in  1927  and  1928  slightly  rose.  It  is,  however,  interesting  to 
note  that,  as  indicated  by  the  curve  at  the  bottom  of  the  chart,  that  slow 
decline  is  not  paralleled  by  clearings,  the  indicators  of  monetary  business 
volume,  and  hence  cannot  have  greatly  affected  business  processes. 

IV.  The  partial  description  of  the  monetary  aspects  of  the  invest- 
ment process  in  the  twenties,  which  is  implicitly  contained  in  the  analysis 
just  presented  (III),  must  now,  in  some  points  at  least,  be  made  explicit 
and  also  supplemented  by  other  data  from  the  financial  sector,  mainly 
the  stock  market. 

a.  For  this  purpose  it  is  first  of  all  necessary  to  insert  into  the  picture 
the  phenomenon  which  so  much  increased  in  importance  during  the  first 
postwar  decade — everywhere,  but  in  particular  in  this  country — Tempo- 
rary Investment.2  We  have  seen  that  industrial  concerns  which  happen 
to  hold  owned  funds3  in  excess  of  what  they  require  at  the  moment — for 
instance,  undivided  profits  or  proceeds  from  the  issue  of  bonds  or  stock — 
insurance  companies,  public  bodies,  wealthy  individuals  in  the  act  of 
shifting  their  investments,  often  find  it  convenient  to  acquire  quick 
assets  of  the  type  that  banks  acquire  for  secondary  reserve  purposes,  or 
even  assets  which  a  bank  could  and  would  not  buy  for  this  or  any  other 
purpose,  and  so  to  take,  for  a  time  or  regularly,  a  hand  in  the  game  of  the 
open  market.  Instead  of  buying  such  things  as  acceptances,  treasury 
bills,  or  even  bonds,  they  may,  of  course,  also  lend  in  the  term's  most 
narrow  sense,  e.g.,  to  bill  brokers  or  stockbrokers.  The  way  in  which 
this  special  case  of  temporary  investment  affects  the  situation  of  banks 
does  not  differ  from  the  way  in  which  it  would  be  affected  by  the  pur- 

1  As  has  been  mentioned  before,  the  English  distinction  between  current  and  deposit 
account  seems  to  the  writer  to  be  somewhat  more  significant  than  the  American  distinction 
between  demand  and  time  deposits.  This  is  why  for  the  curve  at  the  bottom  of  the  chart 
clearings  are  divided  by  current  accounts  only. 

a  The  term  is  to  be  understood  in  the  technical  sense  given  to  it  in  Chaps.  Ill  and  XI. 

8  Though  confining  ourselves  to  owned  funds,  we  should  remember  that  the  phe- 
nomenon itself  is  not  necessarily  confined  to  owned  funds  in  our  sense.  It  should  also 
be  remembered  that  there  is  no  contradiction  in  saying  that,  say,  an  insurance  company 
practices  temporary  investment  permanently. 


1919-1929  871 

chase  of  bonds;  in  fact,  we  may  usefully  call  the  transaction  "buying  a 
loan."  If  the  loan  bought  has  already  been  in  existence  and  "owned" 
by  the  bank  which  holds  the  deposit  to  be  thus  invested,  then  this  bank's 
deposits  and  loans  are  reduced  by  equal  amounts  so  that  the  pressure,  if 
any,  on  its  cash  or  its  reserve  deposit  is,  ceteris  paribus,  relieved  and  its 
lending  power  increased.1  The  banking  system's  aggregate  lending 
power  is  also  increased  and  we  may  say  that  existing  deposits  now  go 
further  or  that  their  "velocity"  rises.  If  the  loan  did  not  exist  before 
but  has  been  newly  created  by  the  transaction,  the  recipient  being  still 
assumed  to  bank  with  the  same  bank  as  the  lender,  the  bank's  deposits 
and  loans  are  not  affected  and  its  lending  power  is  not  increased.2  If 
either  the  purchase  of  the  preexisting  loan  or  the  emergence  of  a  new 
one  leads  to  the  transfer  of  the  corresponding  amount  of  the  investor's 
balance  to  another  bank,  as  in  the  United  States  it  mostly  would,  we 
get  an  additional  feature  which  for  the  investor's  bank  is  in  the  nature 
of  a  shock.  There  are  several  obvious  ways  to  absorb  the  latter,  such 
as  using  elbow  room  the  bank  may  have  been  sensible  enough  to  pre- 
serve, or  borrowing  at  its  reserve  bank  or  from  the  receiving  bank. 
Moreover,  in  the  case  of  actual  transfer  of  reserve  balances,  the  status 
of  the  receiving  bank  will  be  improved  and  its  lending  power  increased, 
normally  by  the  same  amount  by  which  the  paying  bank's  lending  power 
is  curtailed.  But  no  generalization  about  this  can  have  any  realistic 
virtue.  Net  contraction  of  aggregate  loans  in  the  banking  system  is 
perfectly  possible,  especially  if  the  transfer  happens  to  be  from  a  bank 
in  a  strained  position  to  a  bank  in  a  comfortable  one,  though  even  in  this 
case  the  ultimate  effect  for  the  economy  as  a  whole  will  be  much  akin 
to  that  of  an  expansion  of  bank  loans. 

So  far  we  have  dealt  with  temporary  investments  or,  to  take  the 
special  case,  loans  by  nonbanks  within  a  closed  economy.  The  logic  is 
no  different  if  newly  imported  funds  are  used  for  the  purpose,  but  there 
is  a  practical  difference  in  the  complex  of  effects  if  immigration  of  those 
funds  entails  an  addition  to  member  bank  reserves.  Let  us  imagine,  in 
order  to  illustrate  this  point,  that  a  foreigner  (in  the  twenties)  dropped 
from  the  sky  with  the  equivalent  of  a  million  dollars  in  gold  which  he  had 
salvaged  from  oppressive  taxation  or  anticipated  political  trouble  in 

1  Cetera,  however,  need  not  be  paria.     But  assuming  that  the  bank  is  now  able  to 
make  additional  loans,  it  is  worth  while  observing  that,  in  this  as  in  some  other  cases, 
decrease  in  deposits  will  be  accompanied  by  increasing  ease  in  the  money  market  and  a 
tendency  for  money  rates  to  fall.     This  otherwise  trivial  observation  acquires  some  impor- 
tance in  the  presence  of  theories  that  associate  rise  and  fall  in  deposits  with  falling  and 
rising  tendencies  in  interest  rates. 

2  It  might  be  said,  however,  that  it  is  potentially  increased,  if  the  recipient  of  that  loan 
would  otherwise  have  applied  to  the  bank,  a  very  realistic  case :  his  impending  application 
being  warded  off,  the  bank  can  now  view  other  applications  more  favorably. 


872  BUSINESS  CYCLES 

some  other  country,  and  that  he  therewith  acquired  a  balance  in  some 
member  bank.  This,  of  course,  would  swell  that  bank's  reserve  with  its 
Federal  reserve  bank,  and  whatever  the  depositor  does  with  the  balance 
he  acquires — we  may  well  imagine  that  he  will  invest  temporarily  until 
he  has  recovered  from  his  fall — the  effects  of  the  incident  increase  in  bank 
reserves  will  overshadow  any  effects  his  action  may  have. 

The  motives  of  temporary  investors  are  clear.  There  are,  however, 
also  motives  for  them  to  use  their  banks  as  intermediaries — motives  of 
technique  and  others:  the  bank  may,  for  example,  be  a  useful  screen. 
Banks  complied  in  the  first  place  because  they  had  to.  Temporary 
investors  are,  of  course,  competitors  and  their  very  presence  was  a 
symptom  of  the  weakness  of  the  banks'  control  over  the  financial  struc- 
ture. But  in  the  second  place,  a  bank  stood  to  gain  more  than  the 
commission  by  having  the  placement  of  such  funds  left  to  it.  In  par- 
ticular, it  could  then  use  them  in  order  to  relieve  its  individual  position 
and  also  in  order  to  finance  purposes  in  which  it  may  have  been  directly 
or  indirectly  interested  without  caring,  in  times  of  active  criticism,  to 
sponsor  them  too  publicly  suo  nomine.  So  cooperation  was  indicated. 
It  did  not  alter  the  modus  operandi  of  these  loans,  but  it  entailed  the 
statistical  consequence  that  they  were  now  reported  as  Loans  on  Account 
of  Others.  Since  they  were  mostly  directed  toward  the  stock  exchange, 
they  showed  up  as  part  of  Brokers'  Loans,1  while  banks  gave  another 
part  on  their  own  account,  as  a  matter  of  their  business  routine. 

Another  remark  will  complete  the  theory  of  this  special  case  of  a 
special  case  of  Temporary  Investment  and,  incidentally,  our  discussion 
in  Chap.  XIII,  Sec.  D.  In  practice,  a  stockbroker  must  not  only  have 
assets  of  his  own  but  also  keep  a  good  bank  balance.  We  may,  however, 
assume  that  he  will  borrow  part  of  this  balance  against  his  other  assets 
so  that  his  mere  readiness  to  do  business  will  produce  some  brokers'  loans. 
If  speculators  now  do  what  in  Chap.  XIII  we  have  called  "buying  them- 
selves in,"  i.e.,  transfer  to  brokers  certain  sums  to  start  their  margin 
accounts — they  need  not  do  this,  however,  but  can  deposit  securities 
instead — these  sums,  though  they  may  also  be  borrowed,  will  diminish 
the  total  of  brokers'  loans.  The  aggregate  of  the  remainder  will  not 
change  when  brokers  thereupon  get  busy  with  executing  orders  to  buy 
or  to  sell.2  For  though  individual  brokers  will  have  to  borrow  in  order 

1  That  part  includes,  for  any  individual  bank,  also  what  it  thus  lent  on  account  of  other 
(out-of-town)  banks.  The  behavior, of  this  element  was  in  between  the  behavior  of 
brokers'  loans  on  account  of  nonbank  customers  and  of  brokers'  loans  on  the  lending 
banks'  own  account.  Our  argument  primarily  refers  to  the  two  latter  categories. 

*  This  point  is  overlooked  surprisingly  often.  The  writer  is  not  very  familiar  with 
the  literature  of  stock  exchange  operations,  but  as  far  as  he  knows,  credit  for  stressing 
it  is  due  to  Mr.  W.  J.  Eiteman,  Economics  of  Brokers'  Loans,  American  Economic  Review 
for  March  1932. 


1919-1929  873 

to  replenish  their  own  deposits  as  soon  as  they  meet  with  an  adverse 
balance  on  the  day's  transactions,  other  brokers  must,  ipso  facto,  acquire 
a  favorable  balance  which  up  to  the  amount  of  their  indebtedness  will 
reduce  brokers'  loans  by  as  much  as  they  have  been  increased  by  the 
borrowing  of  the  buying  brokers,  unless  it  be  withdrawn  by  customers. 
It  is  true  that  this  will  be  done  very  promptly,  especially  if  the  sellers 
are  not  speculators  but  either  the  issuers  of  the  securities  they  sold  or 
outsiders  without  margin  accounts.  Nevertheless,  it  is  neither  buying 
nor  holding  as  such  that  brokers  have  corporatively  to  finance,  but 
withdrawals  or,  as  we  may  also  put  it,  conversion  of  balances  with  brokers 
into  balances  with  banks.  Hence,  although  brokers'  debts  are  technically 
incurred  in  order  to  service  customers'  debts,  aggregate  brokers'  loans 
can  in  general  only  increase  through  conversion  of  customers'  claims. 

This  conversion  might  create  a  very  awkward  situation.  The  con- 
stant threat  of  it  is,  in  fact,  one  of  the  brakes  that  are  essential  to  the 
normal  working  of  the  money  market  mechanism.  But  if  a  large  amount 
of  owned  balances  stands  ready  to  "buy"  the  claims  against  brokers,1 
new  balances  need  not,  to  that  extent,  be  created,  and  speculators' 
original  deposits  with  brokers  plus  profits  or  minus  loss  can  be  piloted 
into  the  channels  of  consumers'  and  producers'  expenditure  without 
for  the  time  being  causing  any  disturbance — to  the  banks  or  anyone  else — 
except  so  much  as  is  implied  in  hitherto  idle  deposits  becoming  active. 
We  cannot  follow  up  the  interesting  implications  of  this  argument,  but 
it  should  be  observed  that  it  almost  completely  disposes  of  the  theory — 
held  not  only  by  politicians  and  popular  writers  but  apparently  also 
by  some  economists  and  the  Federal  Reserve  Board — that  brokers'  loans 
during  the  twenties  absorbed  funds  that  "legitimate  business"  was 
supposed  to  need  sorely.  On  the  contrary,  brokers'  loans  were  a  means 
of,  as  it  were,  coining  speculators'  gains  and  injecting  them  into  the 
stream  of  economic  life.  They  might,  therefore,  be  called  "inflationary" 
in  an  obvious  sense  of  that  term.  And  an  argument  against  them  can  be 
forged  from  that.2  But  this  does  not  save  the  other. 

It  is  in  this  light,  as  a  concomitant  or  vehicle  of  realized  speculators' 
gains  or  of  stock  issues,  that  we  must  look  upon  brokers'  loans.  They 
increased  strongly  in  1921-1922,  1924-1925,  and  1927-1929.  But  what 

1  For  that  is  what  lending  to  brokers  comes  to.     From  the  standpoint  of  banks,  how- 
ever these  loans  are  newly  created.     It  should  be  borne  in  mind  that  the  argument  applies 
equally,  with  little  modification,  to  the  case  in  which  the  sellers  are  the  issuers  of  the  stock 
as  to  the  case  of  transactions  in  "old"  stocks. 

2  See  Chap.  XIII,  Sec.  D.     It  is,  in  fact,  not  farfetched  to  surmise  that  this  is  what  the 
Federal  Reserve  Board  meant  when  it  tried  to  fight  speculation.     It  may  also  have  resented 
having  its  wires  crossed  by  lenders  beyond  its  control  and  having  to  listen  to  bank  execu- 
tives declaring  in  reply  to  admonitions  that  these  lenders  were  as  independent  of  them 
(the  executives)  as  the  man  in  the  moon,  which,  of  course,  was  not  quite  so. 


874  BUSINESS  CYCLES 

banks  contributed  on  their  own  account  was  not  only  comparatively 
steady — around  1  billion  even  during  the  last  three  years1 — but  actually 
tended  to  decline  through  the  greater  part  of  1928  and,  after  a  strong 
rise  in  the  last  quarter  of  that  year,  again  during  the  first  half  of  1929. 
The  spectacular  ascent  in  1928  and  1929  to  almost  7  billions  was  entirely 
due  to  loans  on  account  of  others,  which  from  the  beginning  of  1928 
to  the  maximum  in  1929  rose  from  about  1  to  nearly  4  billions.2 

b.  We  will  now  inspect  Chart  LII. 

Bearing  in  mind  the  argument  just  presented,  we  shall  not  interpret 
the  covariation  between  brokers'  loans  and  prices  of  industrial  stocks, 
obvious  and  understandable  as  it  is,  in  the  usual  way,  i.e.*  in  the  sense 
that  the  abundance  of  funds  available  for  stock  exchange  operations 
was  the  initiating  force  of  the  boom:  in  part,  especially  in  the  case  of 
foreign  funds,  it  was,  if  anything,  the  consequence  of  it.  The  relation, 
however,  is  more  complicated  than  that.  The  initiating  force  was,  as 
always,  industrial  success.  But  the  abundance  of  money  was  a  condi- 
tion for  speculators'  finding  it  so  easy  to  convert  gains  into  balances  and 
for  concerns'  being  able  to  sell  such  quantities  of  new  securities  at  high 
prices.  Relending  part  of  the  proceeds,  which  in  the  last  years  were 
much  above  immediate  requirements,  in  turn  increased  brokers'  loans. 
This  practice  further  helped  to  relieve  banks  of  the  pressure  which 
conversion  would  otherwise  have  put  upon  them,  and  to  paralyze  for  a 
time  the  brakes  of  the  engine.  Conversion  thus  facilitated  further 
conversion  instead  of  impeding  it,  as  it  would  normally  do,  and  abundance 
of  "funds"  created  additional  abundance. 

We  may  note  that  these  "funds"  would,  on  their  way  to  and  from, 
not  necessarily  but  frequently  pass  through  time  accounts — thus  effecting 
the  utmost  economy  in  the  use  of  demand  deposits — and  that  converted 
funds — or  funds  created  by  conversion — which  were  neither  redirected 
to  brokers  nor  immediately  applied  to  financing  consumption  or  real 
investment  would  regularly  be  placed  on  time  account.  This  is  the  only 
nontrivial  relation  between  time  deposits  and  stock  prices,  i.e.,  the  only 
relation  which  means  more  than  that  both  happened  to  increase  under 
the  influence  of  environmental  conditions  common  to  them.  And  it 
has  nothing  to  do  with  the  investment  of  savings;  hence,  cannot  verify 
any  theories  about  it. 

The  above  analysis  may  be  translated  into  terms  of  money  rates. 
The  theorists  who  establish,  via  the  rates  of  interest,  a  simple  causal 

1  A  sharp  rise,  followed  by  a  still  sharper  fall  within  a  few  weeks,  occurred,  however, 
in  the  fall  of  1929  (see  next  footnote).     The  figures  given  are  those  of  New  York  City 
member  banks,  also  for  the  loans  on  account  of  others. 

2  The  remainder  consisted  of  loans  on  account  of  out-of-town  banks.     The  7  billions 
refer,  it  should  be  remembered,  to  reporting  member  banks.     The  grand  total  was,  for 
September  1929,  about  8.5  billions. 


1919-1929 


875 


1919    1920     1921      1922     1923     1924     1925     1926     1927    1928     1929     1930    1931      1932     1933    1934     1935 
CHART  LII.— United  States  (see  Appendix,  p.  1073). 


876  BUSINESS  CYCLES 

connection  between  abundant  money  and  booming  stock  markets  cannot 
glean  much  comfort  from  the  course  of  events  in  the  twenties.  It  is  true 
that  the  mechanism  described  kept  money  rates  lower  than  they  would 
otherwise  have  been.  But  the  struggle  for  conversion,  though  facilitated 
to  the  utmost,  all  the  same  asserted  itself  in  a  relatively  high  call  rate, 
which  did  not,  however,  spell  tight  money  all  round.  As  to  the  first 
point,  we  observe  that  only  to  the  end  of  the  first  quarter  of  1925  call 
rate  kept  anything  like  its  regular  relation  to  other  open-market  rates, 
especially  the  commercial  paper  rate,  i.e.,  the  relation  which  would 
follow  from  its  place  in  the  structure  of  the  banking  business.  Till 
then  it  was  indeed  (almost)  consistently  lower,  reaching  its  (monthly) 
trough  for  August  1924  when  it  was  2  per  cent,  while  commercial  paper 
rate  still  stood  at  3.25.  But  afterward  it  was  (almost)  consistently  and 
increasingly  higher,  the  difference  reaching  a  maximum  (3.71)  in  March 
1929.  The  presence  of  a  stimulus  to  speculation  from  low  rates  could 
hence — even  if  we  waive  theoretical  objections — be  averred  only  for  the 
second  part  of  1924.  As  to  the  second  point,  open-market  rates,  of 
course,  fluctuated  together.  And,  as  we  have  seen  (Chart  XLIV), 
interest  on  customers'  lines  of  credit  increased  in  1928-1929  by,  on  a 
rough  average,  1  per  cent.  But,  as  we  have  also  seen,  no  restriction 
took  place  in  that  quarter.1  Bond  yield  reacted  but  moderately  (see 
chart).  And  money  raised  by  the  issue  of  stock  was  never  so  plentiful. 
The  "strain"  of  1928  and  1929  was  substantially  confined  to  the  stock 
exchange.  It  came  to  a  head,  as  it  was  bound  to,  through  the  retreat 
rather  than  the  exhaustion  of  the  funds  which  financed  brokers'  loans  and 
which  through  them,  in  the  way  indicated  (see  above,  sub  a),  served  the 
purposes  of  speculators  and  issuers,  while  at  the  same  time  they  were 
by  their  owners  (other  issuers  among  them)  earmarked  and  counted 
upon  for  other  expenditure.  In  the  end,  fright  may  have  been  what 
turned  retreat  into  rout.  And  that  fright  may  have  been  partly  moti- 
vated by  the  perception  of  absurdly  high  stock  prices  and  may  thus  link  up 
with  the  fundamental  fact  that  the  economic  process  was  heading  toward 

1  It  might  be  replied  that  an  increase  in  the  interest  charge  is  restriction.  What 
the  statement  in  the  text  means  is,  first,  that  banks  did  not  restrict  loans  to  business.  As 
has  been  pointed  out,  they  increased  them  and  were  enabled  to  do  so  by  various  favorable 
circumstances  and  by  their  own  action,  i.e.,  by  reducing  their  open-market  commitments. 
Second,  we  hold  that  in  booming  conditions  a  1  per  cent  increase  in  interest  charges  does 
not  induce  significant  restriction  of  business  operation.  Perhaps,  in  strict  logic,  it  should; 
but  it  does  not.  The  economic  world  is  no  billiard  table.  What,  then,  did  that  increase 
mean?  Nothing  except  that  banks  will  increase  their  charges  if  circumstances  give  them 
an  excuse  for  doing  so.  The  monetary  strain  of  1928  and  1929  is  pure  moonshine.  Or 
rather,  he  who  speaks  of  strain  betrays  that,  when  speaking  of  business,  he  is  thinking  of 
margin  accounts.  When  in  1929  the  funds  of  "others"  took  fright,  the  New  York  reporting 
member  banks  stepped  into  the  breach  so  as  to  increase  their  demand  deposits  by  nearly 
1.6  billions.  How  could  that  have  been  done  if  there  had  been  any  strain  beforehand? 


1919-1929  877 

depression.  But  the  situation  in  the  stock  market  was  untenable 
independently  of  all  that.  Precisely  if  business  had  gone  on  booming,  the 
nonbank  lenders  would  themselves  eventually  have  needed  their  funds. 
The  pyramid  of  loans  on  account  of  others,  and  stock  prices  with  it, 
would  have  had  to  come  down  for  this  reason  alone.1 

Some  minor  points  in  the  picture  may  be  noted.  Prices  of  railroad 
stocks  increased  materially  less  than  the  prices  of  industrials.  Yet  the 
writer,  being  unable  to  see  anything  but  gloom  in  the  situation  of  rail- 
roads, finds  it  difficult  to  understand  why  they  should  have  risen  as 
much  as  they  did.  New  York  debits  reflect  the  transactions  of  the  stock 
market  well  though,  owing  to  obviation  and  the  fact  that  its  relative 
importance  is  bound  to  increase  in  consequence  of  an  increase  in  the  total 
of  transactions,  on  a  progressively  reduced  scale.  Temporary  invest- 
ment— brokers'  loans,  in  particular — explains  the  quiet  behavior  of 
New  York  City  bank  loans  and  net  deposits. 

No  great  value  attaches  to  the  data  about  issues  for  the  purpose  of 
securing  new  capital,  embodied  in  Chart  LIII. 

We  may,  however,  note  a  few  significant  features.  There  is,  first,  the 
ominous  increase  in  the  flotations  of  securities  of  investment  trusts  and 
financial  and  trading  companies  since  1926,  which  reflects  the  doings 
of  the  last  three  years  of  the  period  and  teaches  much  about  the  specifi- 
cally American  characteristics  of  the  great  crisis.  The  role  of  the  Land 
and  Buildings  group  should  next  be  noted — much  of  it  was  of  a  similar 
type  and  no  less  indicative  of  troubles  to  come.  Third,  the  industries 
that  carried  the  third  and  fourth  Juglars  are  mostly  represented,  as  we 
should  expect,  though  some  of  them  very  modestly.  Public  utilities 
stand  out  as  the  great  consumer  of  capital.  Fourth,  if  we  discard  the 
classes  that  directly  link  up  with  the  mania,  we  see  the  rise  of  the  fourth 
Juglar  clearly  indicated.  If  it  was  not  more  so,  this  was  due  to  the  fact 
that,  as  stated  before,  concerns  took  advantage  of  the  opportunities  the 
stock  markets  offered  to  them  1928-1929,  even  if  their  immediate 
investment  programs  did  not  require  any  outside  financing.2 

1  The  argument  will  be  developed  from  another  standpoint  in  V.     The  reader  should 
note  both  the  superficial  similarity  and  the  fundamental  difference  between  that  argument 
and  the  Karsten  theory,  see  Journal  of  the  American  Statistical  Association  for  December 
1926.     Also  cf.   Persons  and  Frickey,   Money  and  Security   Prices,  Review  of  Economic 
Statistics  for  January  1926.     Professor  Bresciani-Turroni's  important  contribution  should 
again  be  mentioned:  Considerazioni  sui  Barometri  Economics,  Oiornale  degli  Economists 
for  January,  May,  and  July  1928.     Finally,  comparison  is  invited  with  the  analysis  of 
the  boom  by  Professor  Lindley  M.  Fraser,  American  Economic  Review  for  June  1982. 

2  The  evidence  of  the  chart  may  be  usefully  compared  with  the  increase  in  total  long- 
term  debt.     That  of  railroads  was  11.9  billions  in  1922  and  13.4  in  1930,  that  of  utilities 
was  respectively  8.4  and  14,  industrial  long-term  debt  increased  from  6.8  to  10.8  (U.  S, 
Dept.  of  Commerce,  Long-term  Debt  in  the  United  States,  p.  6). 


878 


BUSINESS  CYCLES 


Millions  of  Dollars 
1,200 


2,100- 


1920 


1,500 


1925 


2^00 


1921 


1922 


1923 


1924 


1926 


1927 


1928 


0  inn 

D    RAILWAYS 

1,800  - 
1,500  - 
1200  - 

D    PUBLIC  UTILITIES 
•                                               i    IRON,  STEEL,COAL,  COPPER 

H    EQUIPMENT  MFGERS. 
8    MOTORS  AND  ACCESSORIES 

:                                               I    LAND,  BLDGS.  ETC 
li    RUBBER 

900- 

•    INVEST.  TRUSTS,TRADING,  ETC 

600- 
300 
0L 

r 

1 

1    1 

; 

Ui 

ill  [ 

ill           I 

1929  1930  1931  1932  1933 

CHART  LIII. — New  capital  issues;  New  York  Stock  Exchange  (see  Appendix,  p.  1074). 


1919-1929 


879 


An  investigation  made  in  order  to  get  at  figures  more  relevant  to  the 
processes  of  real  investment,1  reveals  the  fact  that  in  1929,  out  of  the 
total  of  public  issues  by  domestic  corporations,  which  amounted  to 
nearly  9.4  billions,  at  the  most  %  billions  can  have  been  (directly)  for 
the  purpose  of  providing  real,  or  "  nonfinancial "  or  "productive,"  capital, 


1920  1925  1930 

CHART  LI V.— United  States  (see  Appendix,  p.  1074.) 


1935 


and  even  these  £  billions  include  issues  for  providing  working  capital, 
the  greater  part  of  the  issues  of  installment  finance  and  real  estate  loan 
companies  and  issues  the  purposes  of  which  were  so  ambiguously  stated 
as  to  make  inclusion  or  exclusion  a  mere  tossup.  One  of  the  results, 
however,  was  to  strengthen  confidence  in  the  reliability  of  the  series 

1  It  soon  proved  too  laborious  to  be  extended  beyond  the  year  first  tackled,  1929,  but 
has  been  published  recently.  See  George  A.  Eddy,  Security  Issues  and  Real  Investment 
in  19129,  Review  of  Economic  Statistics  for  May  1937.  Perusal  of  the  appendix  of  that  study 
will  introduce  the  reader  to  the  difficulties  of  the  problem.  Issues  for  the  purpose  of 
repayment  of  bank  loans  have  been  considered  as  "productive"  if  those  loans  themselves 
were  productive  and  not  older  than  three  years. 


880  BUSINESS  CYCLES 

of  "productive"  issues  published  by  Moody's  Investment  Survey.1 
This  has,  accordingly,  been  used  in  Chart  LIV. 

The  order  of  magnitude  of,  and  the  cyclical  fluctuations  in,  these 
"productive"  issues  are  highly  interesting  and  in  themselves  sufficient 
to  dispel  many  errors  about  the  investment  process  of  the  twenties. 
The  annual  figures  rise  from  864  millions  in  1921  to  1,941  millions  in 
1924,  the  maximum  for  the  period  under  survey  (though  it  was  nearly 
reached  again  in  1930).  Then  they  fall,  the  figure  for  1928  being  below 
that  for  1924  by  446  millions.  The  influence  of  the  speculative  excesses 
of  1929  shows,  but  the  figure  for  that  year  is  still  only  1,787  millions. 
This  is  much  more  like  what  we  should  have  expected  from  our  report 
on  the  industrial  processes  of  the  time.  The  perfect  independence 
from  the  course  of  interest  rates,  bond  yields  for  instance,  should  be 
particularly  noticed.  It  would  of  course  be  hazardous  to  aver  that  these 
issues  are  a  perfectly  reliable  indicator  of  investment  in  general — in 
fact,  it  is  certain  that  corporate  issues  will  in  general  differ  in  timing  from 
other  methods  of  financing  real  investment  and  also  from  the  total  of 
actual  expenditure.  There  is  anything  but  satisfactory  covariation 
between  them  and  the  other  series  plotted  on  the  chart. 

c.  Normality  (conformity  to  our  scheme)  is  what  strikes  us  at  first 
glance  when  we  inspect  the  German  chart  (Chart  LV) . 

We  must  make  a  partial  exception  for  Foundations  and  Failures, 
which  increase  and  decrease  in  1928—1929  in  a  manner  not  in  accord  with 
other  indices  or  with  expectations.  But  call  rate,  though,  of  course,  not 
absolutely  lower  than  in  the  United  States,  normalized  its  position  rela- 
tively to  other  rates  in  1925  and  later  on  the  whole  kept  this  place  except 
during  the  first  7  months  of  1927  and  then  again  in  1929.  The  boom 
of  1926-1927  as  reflected  in  stock  prices  and  stock  exchange  transactions 
was  part  of  the  processes  of  the  Juglar  prosperity  though,  within  it,  it 
came  later  than  is  usual.  After  that  we  observe  the  decline,  also  in 
dividends  and  issues,  which  in  a  Juglar  recession  on  a  Kondratieff  depres- 
sion is  the  normal  thing  to  expect.  And  in  fact,  although  there  was 
much  trouble  ahead,  within  as  well  as  without  the  sphere  of  banking,  the 
stock  market  as  such  never  became  a  center  of  difficulties  in  the  subse- 
quent great  depression.  Yet  neither  ease  in  the  open  market  nor  loans 
on  account  of  others  were  absent.  The  influx  of  foreign  credits  created 
conditions  not  dissimilar  to  those  prevailing  in  the  United  States.  The 
ease  from  1925  to  1927  was  closely  linked  up  with  such  loans,  the  pro- 
1  Vol.  25,  No.  86,  October  1933,  p.  1671.  This  series  excludes  additions  to  working 
capital  and  thus,  including  substantially  the  items  plant  and  equipment  only,  really 
serves  our  purpose  better.  With  it,  Moody's  Survey  combines  another,  which  includes 
municipal  and  farm  loan  issues,  which  are  also  considered  as  productive.  The  total  is 
what  has  been  plotted,  but  the  figures  given  in  the  text  refer  to  (domestic)  corporate 
issues  only. 


1919-1929 


881 


ceeds  of  the  issue  of  industrial  bonds  and  stocks — see  chart ;  total  domes- 
tic issues  amounted  to  about  3  billion  reichsmarks  in  the  first  9  months 
of  1926 — being  partly  used  for  the  repayment  of  short  debts  to  banks  or 
for  loans  to  the  market.  Correspondingly,  the  stock  exchange  loans  of 


1924     1925     1926     1927     1928     1929     1930      1931      1932     1933     1934    1935 
CHART  LV. — Germany  (see  Appendix,  p.  1075). 

leading  banks  (Reports  and  Lombards),  after  increasing  rapidly  during 
1926,  reached  a  maximum  of  1  billion  at  the  beginning  of  1927.  But  the 
Reichsbank  which,  throughout  Mr.  Schacht's  tenure  of  office,  ener- 
getically exerted  what  regulating  influence  it  had,  really  meant  business. 
And  the  severe  lesson  administered  to  speculation  in  1927  proved  suffi- 
cient to  stop  the  upward  cumulative  process  and  to  prevent  its  recrudes- 


882  BUSINESS  CYCLES 

cence  for  the  rest  of  the  period,  hence  a  crash  of  the  American  type1  at 
the  end  of  it. 

In  the  London  stock  market  conditions  of  boom  prevailed,  almost 
without  any  setback,  from  the  middle  of  1925  to  the  middle  of  1928, 
after  which  there  were  upper-turning-point  hesitations  for  about  one 
year.  See  Chart  LVI. 

This  boom,  normal  within  the  cyclical  schema  and  also  reflected  in 
"Town"  clearings,  deserves  some  emphasis  because  of  the  divestment 
by  banks  which  accompanied  it  throughout  (see  Chart  LI)  and  because 
of  the  fact  that  it  followed  upon  the  heels  of  the  gold  standard  act.  From 
our  standpoint  there  is  nothing  to  wonder  at  in  this.  As  a  matter  of 
mechanism,  however,  it  should  also  be  observed  that  divestment  corre- 
sponded in  timing  to  the  strong  increase  in  the  London  clearing  banks' 
loans  at  short  notice  to  the  stock  exchange.  Moreover,  although  there 
are  a  number  of  shortest-run  exceptions,  the  inverse  relation  of  the 
movements  in  stock  prices  and  in  call  rate  is  almost  perfect.  The  short- 
money  index2  fell  from  its  maximum  of  163  per  cent  of  the  1913  average 

1  This  is  all  that  the  limitations  of  our  purpose  and  the  writer's  wish  to  deal  more 
adequately  with  American  developments  will  allow  him  to  say  about  the  financial  aspects 
of  the  German  investment  process  and  the  policy  of  the  Reichsbank  during  those  critical 
years.     In  dismissing  the  subject,  he  wishes  to  add  three  remarks.     First,  in  view  of  the 
host  of  abnormalities  in  the  German  situation,  the  normal  complexion  of  the  sector  under 
discussion  must,  in  fact,  be  largely  attributed  to  the  bold  handling  of  the  Reichsbank, 
though  control  was  made  easier  by  the  factors  which  darkened  the  industrial  outlook  after 
1927.     Second,  the  policy  of  the  Reichsbank  no  doubt  presupposed  not  being  afraid  of 
timely  use  of  bank  rate.     But,  partly  because  bank  rate  is  never  the  decisive  element  in 
any  situation,  partly  because  owing  to  the  foreign  credits  it  could  not  have  had  full  effect 
in  this  situation,  such  use  as  was  made  of  it  was  only  of  secondary  importance.     The  latter 
difficulty,  expressing  itself  in  the  unwieldy  mass  of  loans  on  account  of  others,  was  never 
really  overcome,   though  steadfastly  fought.     It  is  highly  instructive  for  anyone,  and 
should  be  even  more  so  to  believers  in  the  key  position  of  interest  rates,  to  recall  that  in 
June  1927  the  first  reaction  of  the  stock  market  to  the  decision  of  the  Reichsbank  to  raise 
its  rate  from  5  to  6  per  cent  was  a  rise  in  stock  prices,  it  being  argued  that  this  would  draw 
foreign  funds:  "stocks  firm  on  dear  money"  was  the  slogan  of  the  hour.     Hence,  third, 
it  was  not  by  raising  bank  rate  but  by  forcing  banks  to  withdraw  what  corresponded  to 
brokers'  loans  on  their  own  account  which  did  the  job.     It  is  true  that  the  measure  took 
effect  so  promptly  because  in  Germany  the  loans  on  account  of  others  were  not,  as  a  year 
later  they  were  in  the  United  States,  almost  exclusively  directed  toward  the  stock  exchange. 
But  it  would  have  been  effective  even  in  this  case.     For  however  little  banks  may  on  their 
own  account  contribute,  that  little  is  of  cardinal  importance  and  its  withdrawal  would  in 
general  be  quite  sufficient  for  any  desired  degree  of  damping.     This  is  not  without  some 
bearing  on  the  responsibility  of  banks  for  booms  not  directly  financed  by  them :  the  officers 
of  a  regular  army  may  be  entirely  innocent  of  the  excesses  committed  by  irregulars  over 
whom  they  really  may  have  no  control;  but  if  they  use  however  small  a  detachment  of 
their  regulars  in  order  to  provide  rallying  points  for  the  irregulars,  they  can  no  longer 
decline  responsibility. 

2  International  Abstract  of  Economic  Statistics,  International  Conference  of  Economic 
Services,  1934,  pp.  105-107. 


1919-1929 


883 


to  48  per  cent  of  it  in  July  1922,  then  rose  and  kept  well  above  that 
average  during  1926.  In  1927  it  fell  below  it,  however,  and  it  was  only  in 
1929  that,  under  the  influence  of  New  York  events,  it  rose  to  anything 
like  abnormal  heights.  The  hesitations  of  the  curve  of  stock  prices 
set  in  while  it  was  still,  falling  and  bank  rate  no  higher  than  it  had  been 


M 


I 


/i 


OCK 


A\/^ 


JON  CALL  RATE  W 


V/ 


EY  A 

O  ST)CK 


\N 


Or 


SHORT  NOTICE 
XCHX  NGE 


A 


V 


\r 


1919    1920   1921     1922    1323    1924    1925    1926   1927    1928    1929  1930    1931    1932    1933    1934   1935 
CHART  LVI.~ England  (see  Appendix,  p.  1075). 

through  the  boom.  In  what  is  an  instructive  contrast  to  the  American 
case  the  subsequent  crash,  or  as  much  of  it  as  was  not  a  repercussion 
from  New  York,  was  more  a  crash  of  unsound  concerns  (Hatry)  than  a 
breakdown  of  a  structure. 

To  the  policy  of  the  Bank  we  shall  return.     But  a  graphic  survey 
of  flotations  may  be  added  here.     The  strict  regulation  of  issues,  which 


884 


BUSINESS  CYCLES 


Millions  of  «£     n 

20 


70- 
60- 
50- 

30- 
20- 


1897 


1903 


1898 


1899 


1900 


1904 


1905 


1906 


1901 


1902 


1907 


1908 


Kf) 

40[~ 
30 
20 
10 

[]  RAILWAYS 

1  MINES 
0  PUBLIC  UTILITIES 

r 

|  SHIPPING 

A 

i  ESTATE  AND  LAND 
i]  i  MOTORS 
^yjlJj     i  RUBBER 
1913            i  OIL 

:;*    -  p 

In   i 

911 

i'. 
1 

Li 

912 

1909                   1910 

10    -=• 


1925  1926  1927  1928  1929  1930 

CHART  LVII. — English  new  capital  issues  (see  Appendix,  p.  1076). 


1919-1929  885 

for  part  of  the  period  amounted  to  prohibition,  must  of  course  be  kept  in 
mind.  In  order  to  show  up  the  postwar  features  of  the  picture  by  con- 
trast, the  survey  has  been  extended  over  the  whole  Kondratieff.  The 
chart  speaks  for  itself  (Chart  LVII). 

The  path  that  leads  from  the  financial  sector  to  real  investment  is 
tortuous  and  unsafe.  Recalling,  however,  what  has  been  said  about  the 
subject  on  various  occasions  earlier  in  this  chapter,  and  once  more 
accepting  Mr.  Colin  Clark's  guidance,  we  cannot  doubt  the  broad  fact 
that  investment  in  the  twenties  was  a  much  smaller  percentage  of  national 
income  than  in  1907. 1  The  expansion  of  production,  which  as  we  have 
seen  was  considerable,  in  fact  took  place  along  with  a  great  relative 
decrease  in  real  as  well  as  monetary  accumulation.  To  conclude  from 
this  that  the  rate  of  economic  growth  does  not  depend  upon  the  rate  of 
accumulation  of  "capital,"  is  not  more  but  also  not  less  reasonable  than 
it  would  be  to  conclude  that  output  does  not  depend  upon  the  degree  to 
which  resources  are  utilized.  We  know  that  change  in  production  func- 
tions is  more  important  than  either.  And  one-variable  relations  do  not 
work  any  better  in  this  case  than  they  do  in  any  others.  But  consider- 
ations such  as  these  are  an  unsafe  basis  for  the  conclusion  which  econo- 
mists of  the  anti-saving  schools  would  evidently  love  to  draw  from  them. 

V.  a.  That  cross  between  European  doctrines  about  banking  and  the 
exigencies  of  the  American  environment  which  is  or  was  the  Federal 
Reserve  System,2  was  from  the  first  endowed  with,  and  in  1917 
acquired  additional,  powers  of  expansion,  little  in  accord  with  those 
doctrines.  Circumstances  too  familiar  to  insist  upon  prevented  them 
from  acting  on  foreign  exchanges,  but  in  all  other  respects  the  "economy  " 
of  reserves  effected3  by  the  act  and  the  amendment  amounted  to  a 

1  National  Income  and  Outlay,  Table  84,  p.  185.     Mr.  Clark's  figures  are  235  million 
pounds  for  1924  and  255  for  1929.     Overseas  investment  fell  below  the  prewar  level  even 
absolutely.     But  his  definition  of  Net  Fixed  Capital  is  much  wider  than  ours.     The  figure 
that  is  relevant  for  us  is  more  akin  to  Mr.  Clark's  Industrial  and  Commercial  Capital 
Outlay,  which  was  respectively  81  and  72  millions  (Table  88,  p.  193). 

2  The  influence  of  a  prominent  banker  whose  thoughts  on  the  subject  were  the  product 
of  German  tradition — which,  of  course,  means  that  in  part  they  were  also  English — is 
clearly  traceable.     There  are  clauses  in  the  act  that  are  suggestive  of  the  old  Reichsbank- 
gesetz — another  instance  of  the  truth  that  nobody  disregards  facts  so  completely  as  the 
theorizing  practitioner,  an  instance  also  for  the  other  truth  that  any  enactment  is  eventually 
adapted  by  the  environment. 

8  The  "economy"  was  brought  about,  first,  by  direct  reduction  of  reserve  requirements, 
second  by  the  concentration  of  member  bank  reserves  at  the  Federal  reserve  banks.  Esti- 
mates of  the  amount  of  the  economy  due  to  direct  reduction  would  vary  somewhat  accord- 
ing to  the  hypotheses  adopted  as  to  the  modus  operandi  of  the  measure.  An  estimate  that 
has  been  widely  accepted,  that  of  the  Federal  Reserve  Bank  of  Richmond,  puts  the  average 
reserve  requirements  of  all  banks  before  the  Federal  Reserve  Act  at  21.09  per  cent  and 
after  the  amendment  of  1917,  at  9.76.  The  special  reduction  for  time  deposits  first  to  5, 
then  to  3  per  cent  acquires  enhanced  importance  if  our  view  of  the  nature  of  time  deposits 


886  BUSINESS  CYCLES 

devaluation  of  the  dollar.  Like  any  devaluation,  it  might  have  remained 
dormant  for  an  indefinite  time.  But  the  war  made  it  effective  in  a  way 
which  should  gladden  the  heart  of  the  quantity  theorists:  from  1914  to 
1920  deposits  in  All  Banks  approximately  doubled — they  increased  from 
about  18.6  billions  to  about  37.7 — and  the  BLS  index  of  wholesale 
prices  also  roughly  doubled  between  1914  and  1919,  which  year,  owing 
to  the  somewhat  speculative  character  of  many  of  the  prices  which  enter 
into  any  wholesale  price  index,  it  is  but  fair  to  choose  for  a  test.1  The 
expansive  power  of  the  system  was  not,  however,  exhausted  thereby. 
As  evidence  from  time  series  already  presented  shows,  it  was  perfectly 
up  to  financing  at  the  new  level  of  prices  and  values  and,  as  has  been 
pointed  out  before,  able  not  only  to  stay  as  expanded  by  the  war  but  also 
to  expand  further.  Two  additional  reasons  for  this  should  be  mentioned 
her<j,, 

First,  the  ratio  between  currency  in  circulation  (outside  the  Treasury 
and  All  Banks)  and  total  deposits  "rose  sharply  in  the  two  war  years, 
but  then  resumed  that  fairly  steady  decline  which  goes  back  at  least  to 
1893,  and  resumed  it  at  nearly  the  same  average  rate  of  relative  decline 
per  year  as  before."2  Evidently  that  process  which  we  have  dubbed 
the  immigration  of  legal  tender  into  banks  and  which  was  so  important 
an  element  of  the  expansion  of  bank  credit  in  the  past,  had  not  been 
completed.  Currency  outside  of  the  Treasury  and  All  Banks,  as  far  as 
the  annual  figures  reported  by  the  Comptroller  of  the  Currency  allow 
us  to  judge,  increased  rapidly  to  a  maximum  for  1920  and  then  fell 

be  accepted.  The  conversion  of  member  bank  reserves  into  deposits  at  the  Federal  reserve 
banks  against  which  the  latter  were  only  required  to  hold  35  per  cent  of  "cash,"  and  replace- 
ment of  gold  certificates  by  Federal  reserve  notes  covered  to  40  per  cent  in  gold  amounted 
to  a  further  reduction  of  reserve  requirements  for  the  banking  system  as  a  whole  and 
multiplied  the  potential  effect  of  the  reduction  of  member  bank  requirements,  thus  provid- 
ing the  machinery  for  "classic"  war  inflation.  The  policy  temporarily  adopted  of  paying 
out  gold  certificates  in  preference  to  notes,  of  course,  mitigated  pro  tanto  the  last-mentioned 
effect. 

1  The  reader  is  presumably  by  now  aware  of  the  fact  that  the  writer  does  not  harbor 
any  too  friendly  feelings  toward  the  quantity  theory  and,  in  particular,  its  modern  revival. 
The  fact  that  it  works  comparatively  well  under  the  circumstances  of  war  inflation  pre- 
cisely shows  that  it  is  worse  than  useless  if  made  to  stand  on  its  own  legs,  i.e.,  on  the  logic 
of  the  equation  of  exchange  as  such.     But  we  must  give  the  devil  his  due  and  cannot  agree 
with  those  economists  who  try  to  deny  or  explain  away  that  fact  itself.     Exact  covariation 
of  prices  and  quantity  of  money  it  is,  of  course,  unreasonable  to  expect.     And  foreign 
demand  for  American  commodities  does  not  provide  independent  explanation  but  was  part 
of  the  very  mechanism  that  the  quantity  theory  assumes  to  exist. 

2  J.  W.  Angell,  Behavior  of  Money,  p.  17.    The  downward  drift,  though  less  pro- 
nounced, is  also  present  in  the  ratio  between  outside  currency  and  adjusted  deposits  subject 
to  check  (ibid.,  p.  19).     Attention  should  be  drawn  to  the  table  at  the  foot  of  p.  17  which 
shows  that  the  English  ratio,  while  more  than  twice  as  high,  also  shows  the  decline  and  at 
approximately  the  same  rate. 


1919-1929  887 

rapidly  to  a  minimum  for  1922,  the  difference  being  roughly  1  billion. 
The  figure  for  1923  shows  a  "  normalizing "  increase,  but  1924  and  1925 
show  decreases  and  1928  and  1929  but  small  increases:  cyclical  drains 
and  reinigrations  are  very  feebly  marked.  We  have  a  seasonally  adjusted 
monthly  series  for  currency  outside  the  Treasury  and  Federal  reserve 
banks  which  moves  better  in  the  cyclical  phases,  indicating  the  rise  of 
the  fourth  Juglar,  falling  in  1927,  keeping  constant  in  1928,  falling  in  the 
first  half  of  1929  and  again,  after  an  understandable  upward  jerk,  toward 
the  end  of  that  year  and  almost  throughout  1930.  This  is  important 
to  keep  in  mind,  considering  the  role  in  the  cyclical  mechanism  attributed 
by  some  theories  to  cash  movements.  For  us  the  point  that  matters  is 
the  accession  of  power  to  expand  which  this  development  brought  to 
member  banks  as  well  as  to  reserve  banks. 

Second,  while  it  is  right  and  proper  to  stress,  as  most  economists  do, 
the  difference  that  exists  between  postwar  and  prewar  gold  standards 
and  monetary  mechanisms  in  general,  it  must  not  be  overstressed.1 
After  all,  during  most  of  the  period  under  discussion,  a  gold  standard 
was  in  force  in  our  three  countries  which  in  important  respects,  though 
not  in  all,  actually  worked  much  as  its  predecessor  had  done  and  was 
managed  in  accordance  with  principles  and  aims  that  did  not  differ 
toto  caelo  from  prewar  practice.  This  will  presently  become  apparent. 
Just  now  it  should  be  noticed  that  in  this  country,  though  it  could  no 
doubt  have  done  very  well  without  any  gold  at  all,  expansion  of  the 
sphere  of  money  and  credit  was  facilitated  by  the  influx  of  gold  (plus 
domestic  production,  which,  including  that  of  the  Philippines,  had  been 
4.9  million  ounces  fine  in  1915,  2.5  millions  in  1923,  and  then  continued 
to  decline  to  2.2  in  1929),  in  spite  of  the  facts  that  the  monetary  gold 
stock  (gold  in  circulation  plus  gold  in  the  treasury  and  the  federal  reserve 
1  In  the  first  draft  of  this  book  it  was  that  difference  that  was  stressed,  but  the  day  of 
that  task  has  passed.  Any  truth  that  fights  against  traditional  habits  of  mind  is  in  danger 
of  being  overstressed.  It  is,  therefore,  not  otiose  to  insist  on  the  following  distinctions. 
First,  if  nothing  had  changed,  either  in  the  monetary  systems  or  in  the  economic  and 
institutional  pattern  of  the  world,  we  might  still  have  acquired  a  different  attitude  toward 
the  gold  standard.  Objectively,  the  gold  standard  of  the  twenties  might  not  have  been 
different,  yet  thinking  might  have  made  it  so.  This  triviality  is  not  superfluous,  because 
some  authors  did,  in  fact,  confuse  new  views  with  new  realities.  Second,  the  identical 
gold  standard  would  have  worked  differently  in  the  postwar  environment,  e.g.,  under  the 
influence  of  rigidities,  barriers,  political  payments,  and  so  on.  This  difference  is  really 
the  most  important  one,  but  whatever  consequences  it  may  have  had  for  the  "classical" 
practice,  it  was  not  in  itself  a  reason  to  abandon  "classical"  theory.  Even  of  "classical" 
practice  stronger  traces  can  be  observed — no  matter  whether  this  was  good  or  bad — if  we 
cease  to  libel  it  by  defining  it  as  a  set  of  rigid  rules  without  relation  to  the  general  state  of  the 
economic  body  and  if  we  disregard  phraseology  (which  however  was  not  very  novel  either) , 
Third,  the  rules  of  the  game  themselves  may  have  been  changed.  They  were  changed, 
but  not  anything  like  as  much  as  is  now  commonly  believed.  The  great  change  came  in 
the  thirties. 


888  BUSINESS  CYCLES 

banks  minus  earmarked  gold)  only  increased  to  1927  (annual  average 
of  daily  figures  for  that  year,  $4.6  billions;  1922,  3.8)  and  that  there 
always  was  "free  gold."1  For  without  that  influx — and  the  free  gold — 
the  Federal  Reserve  System  would  have  had  to  steer  much  closer  to  the 
wind2  and  either  a  problem  of  protecting  the  gold  value  of  the  dollar  would 
have  presented,  or  the  necessity  of  some  devaluation  would  have  imposed 
itself.  As  it  was,  the  abnormal  influx  exerted  its  normal  influence,  which 
was,  to  be  sure,  "managed"  or  regulated  but  not  more  so  than — in 
countries  with  central  banks — it  had  been  in  the  last  prewar*  decades.3 
An  outflow  of  gold  which  under  more  normal  conditions  might  have 
persisted,  turned  into  an  inflow  in  the  last  three  quarters  of  1920.  This 
inflow  lasted  without  interruption  and  in  spite  of  capital  export  until 
December  1924,  and  then  resumed  in  1926  and  again  in  August  1928,  to 
last  until  October  1929,  and  to  set  in  again  in  1930.  Agreement  between 
this  movement  in  the  monetary  gold  stock  and  the  movement  in  total 
deposits  is  not  much  worse  than  it  had  been  of  old.  Persistence  of  "clas- 
sical" relations  is  clearly  suggested. 

b.  The  mechanics  of  what  is  usually  referred  to  as  Federal  Reserve 
Credit  may  be  put  into  two  graphic  nutshells.4  We  will  first  inspect 
Chart  LVIII. 

Reserve  bank  operations,  as  recorded  by  the  Combined  Balance 
Sheet  of  All  Federal  Reserve  Banks,  may  be  summarized  in  terms  of 

1  On  this  and  cognate  points  see  S.  E.  Harris,  Twenty  Years  of  F.  R.  Policy,  in  par- 
ticular, vol.  II,  Appendix  B,  chart  on  p.  761. 

2  Among  other  things  it  could  not  have  so  readily  extended  help  to  foreign  governments 
and  central  banks,  though  most  of  these  transactions,  especially  the  200  million  dollars 
credit  to  the  Bank  of  England  granted  in  the  spring  of  1925,  achieved  their  ends  without 
being  used.     Another  form  of  help  may  be  seen  in  discount  and  open-market  policies 
which  tended  to  stem  the  inflow  of  gold.     This  subject,  which  we  can  barely  touch,  has 
been  ably  discussed  by  A.  Goldstein,  International  Aspects  of  Federal  Reserve  Policy, 
Review  of  Economic  Statistics  for  August  1935. 

3  The  much-discussed  "sterilization"  of  gold,  preventing  "gold  inflation"  during  the 
twenties,  thus  reduces  to  the  fact  that  technically  there  was  always  some  unutilized  margin 
of  gold.     See  below.     The  economists  who  made  so  much  of  that  sterilization  seem  to 
have  had  an  inadequate  idea  of  the  conditions  under  which  the  system  worked  and  which 
made  that  margin  necessary.     Perhaps  they  also  were  victims  of  a  belief  to  the  effect 
that  prices  must  rise  in  a  country  into  which  gold  is  flowing  and  that  only  sterilization  can 
explain  their  failure  to  do  so. 

4  Those  two  graphs  compress,  of  course,  a  large  amount  of  fact.     They  have  been 
prepared  by  Dr.  Carl  E.  Thomas,  who  not  only  worked  up  the  material  but  developed  the 
analysis  of  this  subsection,  much  of  which  is  taken  verbatim  from  his  report.     For  the 
sake  of  convenience,  the  description  (see  also  Appendix)  extends  to  1937.     It  may  seem 
hazardous  to  enter  into  a  subject  to  which  justice  cannot  be  done.     But  as  far  as  it  goes, 
our  argument  is  self-contained.     For  the  rest,  the  reader  is  referred  to  the  well-known 
standard  works  of  Burgess,  Reserve  Banks  and  Money  Market;  Hardy,  Credit  Policies 
of  the  Federal  Reserve  System;  Harris,  Twenty  Years  of  Federal  Reserve  Policy;  and 
Riefler,  Money  Rates  and  Money  Markets. 


1919-1929 


889 


RESERVE  ACCOUNT  OF  MEMBER  BANKS  CALCULATED 
1*2*3*4-5*6-7 


1919   1920  1921    1922   1923  1924   1925    1926  1927  1928   1929  1950  1931    1932  ISB  1934  1935  I9J6 
CHART  LVIII.— (See  Appendix,  p.  1076). 


890  BUSINESS  CYCLES 

eight  accounts  and  of  certain  statistical  relations  which  appear  to  exist 
between  them.1     These  accounts  are: 

I.  Total  Reserves: 

Gold,  United  States  Gold  Certificates  (a  small  amount),  Other  Cash 
(including  silver,  a  small  amount). 

Gold  dominates  this  account.  In  1919,  gold  constituted  approxi- 
mately 97  per  cent,  and  in  1926  approximately  95  per  cent  of  the  total.2 

II.  United  States  Securities: 

This  account  records  changes  in  the  holdings  of  Government  Securi- 
ties, i.e.,  the  bulk  of  Open-market  Operations. 

III.  Bills  Bought: 

This  account  records  bills  acquired  by  the  reserve  banks.  Bankers* 
acceptances  are  included  under  this  title.  During  the  period  under 
review,  bills  were  often  purchased  or  sold  to  offset  seasonal  variations 
in  the  demand  for  currency  and  other  central-market  disturbances  of  a 
routine  nature,  such  as  government  fiscal  operations.  To  this  extent 
these  transactions  are  properly  looked  upon  as  open-market  operations. 
But  they  really  constitute  a  case  intermediate  between  II  and  VI,  because 
the  initiative  often  came  from  the  member  banks. 

IV.  Float.3 

V.  Notes  in  Circulation: 

Federal  reserve  notes  and  Federal  reserve  bank  notes  issued.  This 
includes  Notes  of  Other  Reserve  Banks,  the  amount  of  which  is,  however, 

1  Account  totals  are  taken  as  of  the  "end  of  the  month"  in  all  cases.  When  figures 
are  actually  reported  for  the  last  day  of  the  month,  these  have  been  used.  When  reports 
are  available  as  of  the  end  of  each  week,  the  report  of  the  last  week  in  the  month  has  been 
used.  Although  "average  daily"  figures  are  available  for  all  the  major  accounts,  these 
data  could  not  be  used,  since  they  are  not  in  sufficient  detail  to  enable  checks  upon  content. 

2  Prior  to  January  1984,  gold  was  valued,  in  the  Combined  Balance  Sheet,  at  $20.67  per 
fine  ounce.  After  that  date  it  was  valued  at  $35  per  ounce.  The  gold  account  and  the  gold 
itself  was  then,  at  the  old  value  of  $20.67,  transferred  to  the  Treasury.  In  its  place 
appeared  a  new  account:  "Due  from  the  U.  S.  Treasury."  Changes  in  this  account,  after 
January  1934,  are  recorded  at  the  rate  of  $35  per  ounce. 

8  Before  January  1926,  this  account  is  represented  by  the  difference  between  Uncol- 
lected  Items  and  Deferred  Availability.  Its  movements  were  insignificant  after  1921. 
After  December  1925,  equal  and  offsetting  amounts  of  "Uncollected  Items"  and  "Deferred 
Availability"  appear  upon  the  Reserve  Bank  Statement.  The  previous  difference  is 
recorded  in  two  separate  accounts,  namely,  Float  and  Federal  Reserve  Notes  of  Other 
Reserve  Banks.  It  is  clear  that  previously  notes  of  other  reserve  banks  were  treated  as 
an  uncollected  item.  Since  they  form  a  part  of  Notes  in  Circulation,  it  is  necessary  to 
add  this  item  to  the  Float  as  published,  to  obtain  a  figure  comparable  to  the  previously 
recorded  difference  between  Uncollected  Items  and  Deferred  Availability. 


1919-1929  891 

recorded  for  a  period  sufficiently  long  to  warrant  the  assumption  that  it 
is  not  significant,  so  that  the  item  may  be  roughly  equaled  to  notes  in 
outside  circulation. 

VI.  Bills  Discounted: 

Member  banks'  rediscounts  or  indebtedness. 

VII.  Government  Deposits: 

This  account  was  negligible  until  recently.1 

VIII.  Member  Bank  Reserve  Account: 

This  account,  often  considered  by  practical  bankers  as  the  member 
bank  clearing  account  with  the  reserve  banks,  records  the  indebtedness 
of  the  reserve  banks  to  member  banks. 

It  will  be  seen  that  the  algebraic  sum  of  accounts  I- VII  produces  a 
time  countour  which  is  practically  identical  with  that  of  VIII,  the  member 
bank  reserve  account.  The  difference  in  level  between  the  two  represents 
the  net  sum  of  All  Other  Accounts  appearing  on  the  balance  sheet  of  the 
reserve  banks.2  This  sum  exhibits  practically  no  current  fluctuations. 
Its  amount  was  in  the  neighborhood  of  300  millions. 

Now  we  subtract  Notes  in  Circulation  from  Total  Reserves  plus 
United  States  Securities  plus  Bills  Bought  plus  Float,  and  find  an  almost 
perfect  inverse  and  linear  relation  between  the  result  and  Bills  Dis- 
counted. See  Chart  LIX. 

In  a  statistical — and  a  short-time — sense  it  may  be  said  that  members' 
indebtedness  is  largely  "explained"  by  the  net  of  those  Five  Accounts. 
And  so  is,  hence,  the  reserve  account,  which  is  a  function  of  this  net  and 
its  relation  to  rediscounts  or  indebtedness.3  It  follows,  in  confirmation 
of  what  has  been  read  earlier  in  this  chapter,  that  members  in  the  aggre- 
gate did  not  habitually  borrow  in  order  to  expand  their  operations. 
They  borrowed  primarily  in  order  to  avoid  having  to  contract  them, 
i.e.,  in  order  to  replenish  their  reserve  accounts  when  the  net  of  the  Five 

1  In  the  Bulletin  of  the  Reserve  Board,  published  in  January  1936,  it  is  definitely  stated 
that  the  recent  increase  in  Government  Deposits  was  the  result  of  the  purchase  of  Govern- 
ment Bonds  by  Member  Banks  for  which  they  surrendered  Reserve  Balances.     This 
method  of  government  finance  involves  a  departure  in  technique  from  previous  procedure. 

2  All  Other  Accounts  consist  of:  Non-reserve  Cash,  Other  Securities,  Foreign  Loans 
on  Gold,  Due  from  Foreign  Banks,  Bank  Premises,  All  Other  Resources,  5  per  cent  Redemp- 
tion Fund,  Foreign  Bank  Deposits,  Other  Deposits,  Capital  Paid  in,  Surplus,  All  Other 
Liabilities,  Reserve  for  Taxes. 

3  A  qualification  is  necessary  with  respect  to  government  deposits.     When  they  increase 
at  the  expense  of  members'  reserves  which,  as  pointed  out  above,  has  been  sometimes  the 
case  in  the  last  years,  such  increases  would  have  to  be  added  to  the  latter,  in  order  to  get 
the  right  contour.     For  the  period  under  survey,  however,  this  has  no  importance. 


892 


BUSINESS  CYCLES 


A -TOTAL   RESERVES 

fl»A  PLUS  US.  SECURITfES 

C-B  PLUS  BILLS  BOUGHT 

0-C   PLUS  FLOAT 

E=  D  MINUS  NOTES  IN  CIRCULATION 


1919  1920  1921   1922  1923   1924   1925  1926  1927  1928  1929  1930  1931  1932  1933  1934  1935  1936 
CHART  LIX. — (See  Appendix,  p.  1077.) 


1919-1929  893 

Accounts  decreased.  And  when  that  net  increased,  that  is  to  say,  when 
they  experienced  an  access  of  funds,  they  did  not  primarily  expand 
operations.  They  first  of  all  reduced  their  debt.  In  some  cases  com- 
pensation was  exact,  or  almost  so.  From  November  1924  to  March  of 
1925,  for  instance,  the  federal  reserve  banks  sold  260  millions'  worth  of 
governments  and  bought  75  millions*  worth  of  acceptances.  The  differ- 
ence was  practically  made  up  by  members'  borrowing,  which  amounted 
to  roughly  175  millions.  In  other  cases  the  compensation  was  not  nearly 
exact.  We  are  not  holding  that  expansion  of  loans  and  deposits  was  at 
no  time  accompanied  by  an  uncompensated  surplus  of  borrowing,  still 
less  that  increase  in  our  net  was  wholly  compensated  by  repayment. 
The  rediscounts,  nevertheless,  stand  out  as  the  variable  primarily 
affected  by  the  net  and  very  little  affected  by  anything  else — a  fact 
which  is  very  important  for  any  theory  of  the  American  central  market. 

Two  other  relations  have  been  exhibited  in  the  chart  which  illustrate 
what  fundamentally  is  the  same  point,  and  characterize  the  same  mecha- 
nism. First,  the  close  short-run  covariation,  but  slightly  disturbed  in 
1929,  of  open-market  rates,  represented  by  the  commercial  paper  rate, 
and  rediscounts.  Nothing  can  be  more  indicative  of  member  banks' 
reluctance  to  being  in  debt  than  this  fact,  which  is  due  to  the  promptness 
noticed  before  with  which  they  contracted  open-market  commitments 
in  that  case.  Or,  to  put  it  the  other  way  round,  the  inverse  relation 
between  open-market  rates  and  the  net  of  the  Five  Accounts  indicates 
practically  complete  short-run  dependence  of  the  former  on  the  ebb  and 
flow  of  "funds":  as  a  matter  of  surface  mechanism  and  in  the  short  run, 
variations  in  these  "funds,"  consisting  in  member  banks'  cash  plus 
reserve  balances,  were  the  dominant  factor  in  the  variations  of  their 
debts  to  reserve  banks,  and  variations  in  these  debts  the  dominant  factor 
in  the  variations  of  interest  rates  which,  and  practically  nothing  else, 
were,  hence,  amenable  to  control  by  the  regulation  of  that  ebb  and  flow. 
Now,  putting  aside  internal  drains  and  remigrations  of  cash,  which,  as 
we  have  seen,  were  of  but  secondary  importance  during  the  period  under 
survey  (at  least  after  1923),  we  may  say  that  the  variations  of  members' 
"funds"  were  a  function  of  central  creation  and  of  gold  movements. 
Central  creation  proceeded  from  the  initiative  of  members  (rediscounts; 
hereafter  called  Responsive  Central  Creation)  or  from  the  initiative  of 
federal  reserve  banks  (open-market  operations  in  governments  and  to 
some  extent  also  in  acceptances,  though  these  constitute  an  intermediate 
case — see  above — hereafter  called  Autonomous  Central  Creation),  only 
the  latter  being  now  included  in  that  ebb  and  flow. 

Finally,  gold  movements,  whatever  their  cause,  acted  in  the  same  way 
as  autonomous  central  creation  and  must  hence  be,  in  this  connection, 
algebraically  added  to  it.  But  if  we  take  them  separately  and  if  we 


894  BUSINESS  CYCLES 

choose  intervals  of  time  short  enough,  we  find  that  they  are  positively 
related  to  open-market  rates  with  a  decreasing  lag.1  On  the  chart, 
short-run  gold  movements  have  been  represented  by  month-to-month 
changes  in  total  reserves  (which  were  mainly  gold — see  above),  and 
open-market  rates,  by  the  yield  on  government  certificates  (but  any 
open-market  rate  would  do).  This  completes  our  picture  by  suggesting 
that  total  reserves,  one  of  the  chief  determinants  of  short-run  variations 
in  open-market  rates,  are  in  turn  themselves  influenced  by  the  latter, 
and  thus  supplying  a  link  in  a  general  relation  of  interdependence. 

Very  familiar  traits  (see  Chap.  XIII)  of  prewar  patterns  thus  emerge 
in  the  postwar  picture.  Whatever  theories  we  may  make  our  own,  facts 
behaved  and  mechanisms  worked,  during  the  period  under  survey,  in  a 
manner  that  was  anything  but  revolutionary. 

c.  It  remains  to  sum  up  the  teaching  of  our  charts  from  the  standpoint 
of  the  Federal  Reserve  System  and  to  appraise  the  effects  of  its  policy 
on  the  course  of  cyclical  phases.  To  formulate  this  policy  in  terms  of 
intentions  would  be  an  impossible  task;  for,  though  this  method  is  in 
any  case  hazardous,  it  becomes  impracticable  in  the  case  of  so  acephalous 
a  body  as  the  one  before  us.  A  central  organ  of  banking  would,  under 
the  conditions  and  in  the  mentality  of  this  country,  always  have  an  uphill 
fight  against  the  public,  the  business  community,  speculators,  senatorial 
farmers'  friends,  member  banks,  and  hence  find  it  very  hard  to  acquire 
enough  authority  to  pursue  any  policy  consistently.  But  what  or  who  was 
the  central  organ  ?  There  were,  first,  the  reserve  banks  themselves,  with 
which  from  the  outset  a  considerable  amount  of  initiative  and  autonomy 
had  been  lodged  by  the  Federal  Reserve  Act,  but  which  soon  acquired 
a  sort  of  customary  right  to  initiate  or  "  propose  "  the  general  policies  of 
the  system.  Second,  the  Federal  Reserve  Bank  of  New  York  must  be 
listed  as  a  special  "reserve  authority,"  because  its  prominent  position, 
its  European  connections,  and  its  particular  interests  gave  it  a  power 
and  a  slant  quite  out  of  line  with  those  of  the  others,  and  because  it 
evidently  aspired  to,  and  to  some  extent  conquered,  under  the  leadership 
of  a  strong  man,  the  role  of  "the"  central  bank.  Intimately  connected 
with  the  New  York  bank  was,  third,  the  committee  on  open-market 
operations,  which  eventually  developed  into  the  most  active  element  of 
the  organism.  Fourth,  there  was  the  Treasury,  which  had  by  no  means 
forgotten  its  ancient  role  in  the  money  market.  And,  fifth,  there  was 
the  Board,  which,  sometimes  at  war  with  one  or  more  of  those  other 
organs  of  general  banking  policy,  at  first  tried  to  assert  such  powers  as 
had  been  vested  in  it  but  soon  drifted  into  the  position  of  a  coordinating 

1  The  lag  seems  to  have  stabilized  itself  at  about  4  months.  After  1930  the  relation 
ceases  to  hold.  But  many  factors  in  that  disturbed  time — •"  capital  flight"  from  Europe, 
for  instance — readily  supply  explanation. 


1919-1929  895 

agency.  This  role,  however,  it  filled  with  astonishing  success,  in  spite 
of  the  fact  (or  because  of  it?)  that  there  are  unmistakable  symptoms 
pointing  to  a  serious  division  of  opinions  within  it,  which  often  paralyzed 
decision.  The  writer  understands  that  election  to  a  seat  on  the  board 
was  a  coveted  honor.  But  the  seats  must  have  been  extremely  uncom- 
fortable ones,  even  if  economists  of  all  nations  had  not  completed  vexa- 
tion by  discovering  that  the  Federal  Reserve  System  was  an  entirely 
new  departure,  harboring  unheard-of  possibilities,  sure  to  put  an  end  to 
the  recurrence  of  depressions,  etc.,  and  by  thus  raising,  even  among  other- 
wise sane  and  intelligent  people,  hopes  that  were  as  unwarranted  as  their 
disappointment  was  certain  to  be  attributed,  in  self-defense  of  the 
enthusiasts  of  scientific  and  extrascientific  description,  to  the  incompe- 
tence of  the  board.1 

But  fortunately  we  need  not  trouble  about  intentions  and  phraseo- 
logies. The  actual  behavior  which  the  logic  of  the  situation  soon  shaped 
into  a  definite  pattern — 1922—1923  roughly  dates  the  decisive  steps  in 
its  evolution — suffices  for  our  purpose.  Before  that  date,  a  tendency 
to  help  in  the  process  of  normalization  after  the  war,  as  evidenced,  for 
example,  by  the  circular  which  reminded  banks  that  war  credits  would 
not  indefinitely  be  prolonged,  is  the  only  thing  to  note.  Criticism  of  this 
attitude  and  the  discovery  that  those  outlets  for  the  reserve  "funds," 
which  the  Federal  Reserve  Act  had  contemplated,  would  not  develop — 
i.e.,  the  business  man's  typical  mortification  at  the  absence  of  adequate 
earnings — perhaps  provided  the  first  impulse  to  embark  upon  buying 
governments  at  the  beginning  of  1922. 

Turning  again  to  Chart  LIX,  we  are  immediately  struck  by  the  fact 
that  no  energetic  use  was  ever  made  of  the  discount  rate.  Not  only  was 
it  always  kept  below  even  the  commercial  paper  rate — thus  never  occupy- 
ing the  position  or  filling  the  function  of  German  or  English  bank  rates — 
but  it  also  followed  the  market  in  every  single  instance  except — and 

1  That  body  was  so  unfortunate  as  to  draw  fire  even  from  the  other  party  to  the 
Battle  of  "Theories."  Economists  who  did  not  share  those  hopes  and  quite  correctly 
appraised  the  limitations  of  the  doctrines  on  which  they  were  based,  are  or  were  in  the 
habit  of  accusing  the  board  of  misdirected  measures  responsible  for  what  most  of  those 
economists  would  call  inflation  and,  partly  or  wholly,  for  the  subsequent  crash.  There 
are  two  mistakes  in  this,  apart  from  the  incorrect  allocation  of  responsibility  to  the  Board. 
First,  such  criticism  seems  to  imply  that  these  economists  yield  to  their  opponents  to  the 
extent  of  sharing  part  of  the  latter's  exaggerated  ideas  of  the  importance  of  what  central 
bank  policy  can  make  or  mar.  Second,  they  do  not  seem  to  take  adequate  account  of  the 
structure  of  the  American  financial  engine  and  of  the  data  confronting  the  board  from 
the  outset.  The  present  writer  is  not  offering  laurels  to  the  board.  If  he  were  hunting  for  the 
most  appropriate  sculptural  ornament  for  the  board's  building,  a  statue  of  Hercules  would 
certainly  not  be  among  the  first  to  occur  to  him.  But  critics  do  make  their  task  a  little 
easy.  It  should  be  added  that  our  list  of  "central  organs"  is  by  no  means  complete;  but  it 
serves  to  convey  the  essential  point. 


BUSINESS  CYCLES 

even  this  exception  is  more  apparent  than  real — at  the  very  beginning 
of  1928,  when  an  effort  was  made  "to  gain  control."  But  at  the  same 
time  it  follows  from  our  analysis  that  there  was  another  reason  for  this 
besides  the  "political  impossibility"  of  handling  that  weapon  boldly:  to 
some  extent  the  Federal  Reserve  System  in  fact  controlled  the  situation 
of  which  market  rates  were  the  outcome  and  symptom,  so  that  in  fol- 
lowing or  ratifying  market  by  bank  rates  it  really  followed  and  ratified 
the  result  of  its  own  action — bank  rate  as  an  independent  weapon  had, 
in  fact,  practically  ceased  to  exist.  Such  leadership  or  management  as 
there  was,  was  exerted  through  the  Five  Accounts,  mainly  by  open- 
market  operations,  supplemented  by  suasion.  As  we  have  seen,  this 
method  requires  in  order  to  be  effective  a  certain  behavior  on  the  part 
of  members,  in  particular  a  certain  attitude  toward  being  indebted 
to  reserve  banks,  for  if  they  had  all  stormed  for  more  credit  and  for 
permission  to  use  it  permanently  and  as  a  matter  of  course,  they  would 
have  had  the  public  behind  them  almost  to  a  man,  and  any  resistance 
by  the  board  or  the  reserve  banks  would  have  been  swept  away.  But 
this  kind  of  discipline  the  better  class  member  banks  displayed  through- 
out, and  the  greatest  achievement  of  the  board  was  its  success  in  training 
them  up  to  it  and  in  establishing  a  professional  tradition  which  made  it 
derogatory  to  a  bank's  standing  to  be  in  the  red  at  its  reserve  correspond- 
ent, except  in  order  to  tide  over  temporary  pressure,  or,  for  a  first-class 
bank,  to  be  in  the  red  at  all,  even  when  the  reserve  system  seemed  to 
invite  it  by  low  rates.  A  qualification  has  to  be  added,  however.  We 
have  said  that  discipline  was  kept  unbroken  until  the  crisis.  This  is 
true,  for  the  whole  period,  only  in  a  formal  sense.  In  the  spirit  it  holds 
only  to  the  beginning  of  1928.  Then  many  important  banks,  especially 
New  York  banks,  kicked  over  the  traces  by  creating  acceptances  and 
selling  them  to  reserve  banks,  not  so  much  in  order  to  evade  the  higher 
discount  rate — this  was  a  by-product — as  in  order  to  evade  the  necessity 
of  running  deeper  into  debt.  The  reserve  banks  bought,  and  were  by 
the  board  allowed  to  buy,  all  that  paper  without  demur.  This  is  the 
most  important  instance  of  a  practice  which  induced  us  above  to  qualify 
our  inclusion  of  purchases  of  acceptances  in  open-market  operations. 

Barring  this,  the  mechanism  functioned.  Although,  as  has  been 
pointed  out,  open-market  operations  did  not  directly  increase  and  decrease 
the  borrowing  facilities  available  to  nonbank  firms  because  the  member 
banks  (largely)  compensated  them — if,  indeed,  they  were  not  from  the 
outset  intended  to  compensate,  hence  to  be  compensated  by,  gold  move- 
ments—  the  policy  was  not  thereby  defeated  since,  whenever  open-market 
operations  decreased  the  net  of  the  Five  Accounts,  banks  were  forced 
to  borrow  or — which  served  still  better — to  reduce  their  open-market 
commitments.  But  the  corollary  to  this  is  that  open-market  sales  never 


1919-1929  897 

went  further  than  this,  t.e.,  beyond  the  compensating  powers  of  banks. 
Perhaps  they  were  not  intended  to.  The  committee  on  open-market 
operations  had  every  reason  to  fear  the  reaction  if  its  measures  had 
touched  "legitimate  business."  In  any  case,  the  effects  which  can 
be  produced  by  this  weapon,  unless  reinforced  by  simultaneous  refusal 
to  buy  acceptances  and  by  rationing  credit  to  member  banks,  are  limited 
to  the  available  ammunition,  which  it  may  be  impossible  to  collect. 
The  reserve  banks'  annual  average  of  United  States  securities  holdings 
in  1919  Was  261  millions — no  great  impression  could  have  been  made 
by  any  sale  from  this  if  there  had  been  any  wish  to  influence  in  this  way 
the  situation  then  prevailing.  Luckily,  the  first  open-market  operation 
to  be  undertaken  with  a  therapeutic  intention  was  a  purchase  and  not  a 
sale:  from  January  to  May  1922,  about  400  millions  worth  of  governments 
was  bought  on  the  top  of  a  steady  influx  of  gold.  In  a  sense  this  was  the 
decisive  step  that  raised  the  Five  Accounts  to  the  level  under  which 
they  were,  except  in  1923,  never  again  allowed  to  fall,  and  thus  per- 
petuated the  plethora  of  money.  The  alternative  would  have  been 
to  enforce  the  liquidation  of  member  bank  indebtedness  without  pro- 
viding the  funds  for  it.  This  would  have  spelled  pressure,  slower 
recovery,  sobered  advance,  much  less  speculation,  and,  after  1929,  a 
milder  depression. 

No  real  pressure  was  exerted  by  the  next  open-market  operation, 
which  was  the  first  to  offset  a  gold  influx  and  consisted  in  the  sale  of 
about  525  millions  United  States  securities  between  June  1922  and  July 
1923.  Selling  actually  was,  while  it  lasted,  at  a  greater  rate  than  gold 
influx,  so  that  member  banks  were  for  the  time  being  forced  to  borrow, 
but  the  continuing  stream  of  gold  alone  soon  made  up  for  this  and 
even  during  that  time  effects  were  amply  compensated  by  gold  plus  bills 
bought  plus  borrowing,  so  that  they  did  not  extend  beyond  the  central 
market.  It  should  also  be  observed,  in  order  to  get  at  a  sound  appraisal 
of  the  achievements  of  this  policy,  that,  as  far  as  the  preceding  open- 
market  purchases  and  the  gold  influx,  as  well  as  the  incident  reduction 
of  member  banks'  indebtedness  by  over  600  millions,  did  speed  up 
recovery — and  while  we  deny  that  it  mode  recovery  and  hold  that  it  was 
all  but  futile  as  a  remedy  for  depression,  we  do  admit  that  the  temperature 
of  positive  phases  can  be  raised  by  central  bank  policies — they  also  helped 
to  bring  about  the  situation  which  the  board  or  the  reserve  banks  then 
felt  to  stand  in  need  of  restrictive  correction.  Thus  a  current  view  on 
the  success  of  this  policy  requires  to  be  corrected  on  two  distinct  heads : 
first,  success  in  influencing  cyclical  situations  was  much  smaller  than  it 
looked  to  post  hoc  ergo  propter  hoc  analysis;  second,  such  success  as  there 
was,  in  part  consisted  in  correcting  the  effects  of  the  reserve  system's  own 
policy, 


898  BUSINESS  CYCLES 

What  we  might  call  the  first  conjuring  trick  followed.  In  the  Kitchin 
recession  and  depression  of  1928—1924,  when  finance  and  business  felt 
a  vague  malaise  after  their  doings  in  1922-1923,  the  reserve  system  stepped 
forward  to  chase  away  darkness,  to  insure  stability  of  prices  and  to 
guarantee  new-era  prosperity  by  buying  about  510  millions'  worth  of 
governments  between  December  1923  and  September  1924,  which,  in  the 
minds  of  some  economists,  was  simply  equivalent  to  increasing  active1 
deposits  by  5,100  millions.  In  reality,  it  did  not  mean  this  but  deposits 
did  react,  as  well  as  member  banks'  debts,  and  the  incident  fall  in  open- 
market  and  reserve  bank  rates  turned  the  tide  of  gold.  The  latter 
effect  was  really  the  most  important  part  of  the  success,  inasmuch  as, 
partly  because  of  European  complaints  and  partly  because  of  a  percep- 
tion of  the  difficulties  which  continued  influx  of  gold  might  eventually 
produce,  the  turning  of  the  tide  was  one  of  the  objects  of  the  measure.2 
As  soon  as  November  1924,  selling  was  again  resorted  to  and  continued 
to  March  1925,  when  the  total  amount  reached  was  260  millions.  This 
operation  has  already  been  mentioned  in  order  to  illustrate  the  modus 
operandi.  Effects  were  as  before,  but  the  operation  deserves  to  be 
emphasized  as  a  token  and  measure  of  the  success  with  which,  braving 
many  difficulties,  the  reserve  system  kept  its  hand  on  the  steering  wheel 
and  actually  followed  a  definite  course.  And  the  same  may  be  averred 
about  the  two  smaller  operations,  purchases  in  the  amount  of  65  millions 
in  April  1926,  and  sales  in  the  amount  of  80  millions  in  August  and 
September  of  the  same  year. 

A  situation  closely  similar  to  that  of  1923-1924 — the  similarity 
extends  to  the  presence  of  European  influence,  this  time  accentuated  by 
a  pilgrimage  to  the  miraculous  shrine — induced  another  purchasing 
campaign  in  May  1927  which  lasted  to  December  and  raised  the  United 
States  securities  holdings  of  the  reserve  system  by  about  300  millions  or 
to  about  double  their  previous  figure.  There  was  by  then  nothing  unusual 
either  in  the  occasion  or  the  effects  or  the  amount  of  the  operations.  If 
we  did  not  know  the  contrary,  we  should  not  from  the  facts  of  the  case 
infer  that  any  controversy  would  have  arisen  about  this  particular  opera- 
tion or  that  the  latter  would  have  been  considered  as  particularly  admir- 
able by  some  and  as  particularly  mistaken  by  others,  or  finally  that 

1  Time  deposits  in  reporting  member  banks,  however,  increased  during  1924  by  nearly 
%  billion. 

2  As  far  as  it  was,  the  purchasing  operations  may  have  commended  themselves  to 
members  of  the  board  who  neither  believed  in  their  efficacy  as  a  remedy  for  the  relapse 
in  business  nor  would  have  approved  of  using  the  tool  at  that  time,  even  if  they  had  believed 
in  its  efficacy.     Thus,  as  happens  so  frequently  in  all  spheres  of  life,  men  were  able,  from 
entirely  different  premisses  and  with  different  and  conflicting  objects  in  mind,  to  agree  on 
the  particular  measure. 


1919-1929  899 

praise  or  blame  for  it  should  have  been  fastened  upon  a  particular  man.1 
The  reason  for  all  this  is,  of  course,  in  the  fact  that  that  was  the  last 
buying  operation  before  the  stock  exchange  crash  of  1929.  We  fully 
share  this  motive,  since  we  have  had  to  stress  both  in  our  historical  nar- 
rative and  in  our  time-series  analysis  many  an  abnormality2  in  the  eco- 
nomic processes  of  1928  and  1929.  But  no  amount  of  careful  searching 
establishes  the  connection  one  has  a  perfect  right  to  suspect.  Such  a 
search  reveals  indeed  many  individual  facts  which  point  to  the  presence 
of  "artificial  stimulation."  For  instance,  gold  exports  were  more  than 
offset,  and  members'  investments,  total  deposits,  and — what  is  particu- 
larly significant — brokers'  loans  for  own  account  of  member  banks,  not 
only  on  account  of  others,  were  increased.  But  the  effects  of  all  this 
may,  as  we  know,  easily  be  exaggerated,  and  nothing  really  adequate  for 
explaining  either  the 'boom  of  1928-1929  or  the  subsequent^  breakdown 
can  be  linked  to  that  buying  operation  per  se  as  the  sole  or  as  a  major 
cause. 

In  fact,  there  is  good  reason  to  doubt  whether  it  played  any  role  in 
the  causal  pattern,  for  the  reserve  system  acted  with  more  than  usual 
vigor  to  stamp  out  any  sparks  that  its  action  may  have  set  flying  about. 
First,  mere  cessation  of  buying  had,  because  of  the  outflow  of  gold,  some 
restrictive  effect  as  early  as  December  1927.3  And  in  January  1928  the 
reserve  system  embarked  upon  the  biggest  selling  operation  but  one 
of  its  history,  in  which  it  persisted  fully  as  long  as  was  compatible  with 
safety,  i.e.,  to  April  1929,  when  sales  summed  up  to  405  millions  and 
total  open-market  operations  since  January  1922,  as  already  stated,  to 
minus  65  millions.  Simultaneously  bank  rates  were  repeatedly  raised, 
eventually  (July  1928)  to  5  per  cent,4  member  bank's  indebtedness  (which 
however,  as  we  have  seen  above,  did  not  in  this  case  quite  fulfill  its 
function)  passed  the  billion  line,  below  which  it  had  moved  ever  since 
1922,5  and  brokers'  loans  for  own  account  dropped  below  1  billion,  while 
gold  exports  continued  through  July.  This  seems  to  have  been  drastic 
action  if  ever  central  bank  action  was.  Nor  was  it  futile,  as  some,  or 

1  Mr.  B.  Strong's  influence  may  have  been  much  in  evidence  on  this  occasion.     But 
the  board  certainly  did  not  offer  significant  resistance,  for  thtey  obligingly  brought  the 
reserve  bank  of  Chicago  into  line  when  it  displayed  reluctance  to  reduce  its  rate  along  with 
the  others.     Nothing  would  have  been  easier  than  to  checkmate  Mr.  Strong  by  letting 
Chicago  have  its  way,  which  would  have  tightened  the  New  York  market. 

2  They  cannot  be  listed  again,  but  if  the  reader  wishes  to  get  the  argument  in  full  he 
should  now  go  back  and  list  them  for  himself. 

3  It  is,  however,  true  that  the  Treasury's  needs  for  funds  arising  out  of  the  conversion 
of  the  second  Liberty  Loan,  which  induced  it  to  take  and  partly  use  an  overdraft  at  the 
Federal  reserve  banks,  may  have  counterbalanced  that  effect  temporarily.     But  the  net 
efflux  of  gold  was  67.42  millions  in  December  alone. 

4  The  last  step,  from  4.5  to  5,  was,  however,  taken  by  only  eight  reserve  banks. 

6  The  monthly  average  for  1920  had  been  nearly  2.6  billions  and  that  for  1921  over  1.7. 


900  BUSINESS  CYCLES 

oppressive,  as  other  people  thought.  It  tightened  the  central  market. 
It  did  not  exert  pressure  on  industrial  and  commercial  business,  as  is 
clear  from  the  simultaneous  increase  in  customers'  loans  which  has  been 
noticed  before.  And  it  was  quite  rationally  discontinued  in  the  second 
half  of  1928,  when  the  gold  influx,  which  resumed  in  August,  was  allowed 
to  take  effect. 

If  the  board  needed  any  lesson,  it  was  supplied  by  this  experience, 
which  conclusively  showed  that  (relatively)  high  rates  will  attract  gold, 
that  for  this  and  other  reasons  no  impression  could,  by  ordinary  methods 
applied  within  ordinary  limits,  be  made — except  very  temporarily  (middle 
of  1928) — on  the  stock  exchange,  and  that  the  most  immediately  danger- 
ous breach  in  the  wall  was  the  loans  on  account  of  others.  It  is  sub- 
mitted that  the  inferences  which  might  have  been  drawn  from  this  lesson 
are,  first,  that  bank  rates  should  be  lowered1  and,  second,  that  loans 
to  the  stock  exchange  should  be  attacked  directly,  though  the  lowering 
of  bank  rates  through  turning,  stopping,  or  reducing  the  flow  of  gold 
would  already  have  done  part  of  the  job.  The  first  was  not  done,  per- 
haps because  it  seemed  too  unorthodox  a  thing  to  do.  But  the  second 
was  exactly  what,  obviously  taking  its  courage  in  both  hands  and  drop- 
ping its  attitude  of  dignified  reserve,  the  board  actually  did2 — or  tried 
to  do.  It  announced  its  intention,  in  the  famous  circular  that  was 
published  on  Feb.  7,  1929,  "to  restrain  the  use,  either  directly  or  indi- 
rectly, of  federal  reserve  credit  facilities  in  aid  of  the  growth  of  specula- 
tive credit/'  The  announcement  was  embarrassed,  overcautious,  and 
also  not  quite  sound  in  explicit  and  implied  argument,  but  in  essence  it 
proclaimed  Mr.  Schacht's  policy  of  1927,  i.e.,  the  policy  of  forcing  banks 
to  withdraw  their  loans  to  brokers,  whereby  other  lenders  would  be 
induced  to  retreat.  No  doubt  is  possible  that  this  measure  would 
have  been  quite  successful  if  resorted  to  earlier,  say,  by  one  or,  still  better, 
two  years.  Then  it  would  have  prevented,  whereas  at  the  beginning  of 
1929  it  could  only  have  precipitated,  the  crash.  The  perception  of  this 
accounts  both  for  the  determined  opposition — led  by  the  reserve  bank  of 
New  York3 — and  for  the  readiness  of  the  board  to  yield,  in  June,  after  its 

1  We  thus  find  ourselves  pro  tanto  in  agreement  with  many  fellow  economists,  among 
them  Professor  Irving  Fisher  and  Dr.  Currie,  with  whose  arguments  and  results  we  do  not 
otherwise  agree.     It  will,  however,  be  realized  that  we  arrive  at  the  same  result  from  very 
different  premisses  and  expecting  quite  different  consequences. 

2  There  are  many  methods  for  achieving  the  same  result,  taxation  amounting  to  con- 
fiscation of  gains  from  stock  speculation  and  others,  but  the  one  chosen  was  presumably 
the  only  one  open  to  the  board. 

3  That  bank  followed  up  its  previous  performance  in  the  field  of  central  bank  policy 
by  the  proposal,  made  a  few  days  after  the  board's  announcement  of  direct  action,  to 
raise  the  rediscount  rate  instead.     If  the  above  analysis  is  correct,  this,  through  attracting 
additional  gold,  would  have  given  another  impulse  to  speculation      The  reader  will  recall 
a  similar  observation  previously  made  in  our  discussion  of  a  similar  situation  in  Germany. 


1919-1929  901 

policy  had  fully  demonstrated  its  effectiveness  by  putting  a  stop  to  the 
rise  in  stock  prices  and  reducing  brokers'  loans  by  between  600  and 
700  millions.  By  then,  however,  it  mattered  little  whether  it  yielded 
or  not. 

The  reader  will  please  formulate  for  himself  whatever  may  seem  to 
him  to  follow  from  this  analysis  as  to  the  merits  of  the  actual  practice 
described  or  of  the  various  alternatives  advocated  at  the  time  and  later. 
For  the  purposes  of  this  book  it  is  sufficient  to  state  or  restate  three 
conclusions  to  which  we  have  been  working  up  all  along. 

First,  whatever  conflict  of  "theories"  and  intentions  there  may  have 
been,  we  have  seen  that  the  actual  behavior  of  the  organs  of  central 
banking  in  this  country  reveals  a  very  definite  pattern.  This  pattern 
greatly  differs,  of  course,  from  that  which  we  have  tried  to  piece  together 
for  prewar  times  from  the  behavior  of  such  central  organs  as  the 
American  banking  system  then  possessed,  since  these  had  little  facility 
for  concerted  action.  But  it  differs  much  less,  in  essentials  as  dis- 
tinguished from  forms  and  methods,  from  the  prewar  patterns  of  central 
banking  in  countries  which  then  had  individual  central  banks.  Gold 
movements,  in  particular,  and  the  state  of  gold  reserves  indeed  did  not 
play  the  role  which  it  was  usual  to  assign  to  them  in  prewar  theories  of 
central  banking.  But  this  was  partly  due  to  exceptionally  favorable 
circumstances  which  would  also  have  pushed  reserve  considerations  into 
the  background  for  any  prewar  central  bank;  and  though  the  effects  of 
gold  movements  were  no  doubt  regulated,  as  they  had  always  been,  they 
were  not  therefore  obliterated.  Open-market  operations  were  no 
novelty,  nor  did  they  serve  in  fundamentally  new  ways  or  for  fundamen- 
tally new  purposes.  American  conditions  put  them  into  the  limelight 
and  enforced  their  systematic  use,  but  that  was  all. 

Second,  it  follows  from  this,  together  with  our  general  view  on  the 
role  and  possibilities  of  central  banking,  that  the  cyclical  processes  of  the 
period  were  not  substantially  affected  by  the  policy  of  the  reserve  system. 
Its  attitude  to  the  ups  and  downs  of  business  was  much  the  same  as  the 
attitudes  of  central  banks  had  been  for  decades  before  the  war.  It 
"regulated"  the  central  market  and,  doing  this  ably  and  conscientiously, 
achieved  what  can,  and  failed  to  achieve  what  cannot,  be  achieved  in  this 
way  and  what  equally  ably  and  conscientiously  managed  central  banks 
achieved  and  failed  to  achieve  before.  The  successes  of  1924  and  1927 
and  the  "failure  to  prevent  the  great  depression"  are  largely  figments  of 
the  mistaken  theories  of  the  day.  With  respect  to  that  failure,  however, 
a  secondary  though  still  important  qualification  is  necessary. 

Third,  this  qualification  bears  upon  the  well-known  controversy 
whether  or  not  there  was  "inflation"  or  "deflation"  and  whether  or  not 
the  reserve  system  and  its  policies  were  responsible  for  it.  Couched  in 


902  BUSINESS  CYCLES 

less  ambiguous  terms,  the  question  reads  whether  there  are  in  the  behavior 
of  time  series  any  deviations  from  expectation  which  are  traceable  to  acts 
or  attitudes  of  the  reserve  system  or  other  autonomous  monetary  factors. 
All  the  evidence  presented  points  to  the  conclusion  that,  so  far  as  fluctua- 
tions and  "trends"  were  concerned,  this  was  not  the  case  until  the  spring 
of  1928.  Quantities,  prices,  values,  incomes,  and  so  on  behaved  sub- 
stantially as  we  should  have  expected  them  to  behave  in  the  absence  of 
disturbance  from  the  monetary  sphere,  although  monetary  magnitudes 
and  expressions  moved  on  a  level  explainable  only  by  the  uifcorrected 
monetary  disturbance  caused  by  the  war  and  by  war  finance.  The 
attempts  that  have  been  made  to  allocate  responsibility  for  the  occur- 
rence of  the  great  depression  to  monetary  and  banking  factors,  or  to  the 
behavior  of  economic  elements  primarily  shaped  by  them,  sound  as 
unconvincing  if  they  come  from  authors  with  "inflationist,"  as  they  do 
if  they  come  from  authors  with  "deflationist"  sympathies.  In  particu- 
lar, there  is  little  more  than  assertion  in  the  theories  which  hold  that 
interest  rates  or  prices  were  either  too  high  or  too  low  or  declining  too 
slowly  or  too  rapidly.  Our  methods  are  rough.  We  must,  hence,  admit 
the  possibility  and  even  likelihood  that  there  may  be  elements  of  truth 
in  all  those  and  similar  assertions,  even  in  mutually  contradictory  ones. 
But  they  are  obviously  inadequate  to  the  phenomenon  to  be  explained: 
common  sense  and  common  experience  should  be  sufficient  to  convince  us 
that  if  explained  by  maladjustments  such  as  these,  the  great  depression 
remains  unexplained.  Reserve  policy  in  particular  seems  rather  to  have 
warded  off  than  to  have  caused  disturbance,  rather  to  have  eased  than 
to  have  impeded  adaptation,  rather  to  have  brought  out  than  to  have 
destroyed  or  controlled  the  essential  (expected)  contours  of  the  cyclical 
process — the  last  point  being  illustrated,  for  instance,  by  the  fact  that 
rediscounts  moved  so  well  in  the  cyclical  phases.  The  paradox  that  has 
been  so  keenly  felt  by  economists  and  noneconomists,  viz.,  how  such  a 
thing  as  the  great  depression  could  have  occurred  if,  though  not  every- 
thing, yet  so  much  in  the  economic  organism  was  "quite  all  right," 
obviously  calls  for,  and  is  resolved  by,  the  more  fundamental  explanation 
which  the  logic  of  the  capitalist  process  supplies. 

But  if  that  is  so  with  respect  to  the  occurrence  of  the  depression,  it  is 
not  necessarily  so  with  respect  to  its  intensity.  And  the  latter  links 
up  with  the  breakdown  of  stock  markets  and  this  again  with  the  abnor- 
malities of  the  last  2  years  of  the  period.  Then  it  was  that  things  got 
out  of  hand,  in  spite  of  the  fact  that  the  reserve  system  used  its  regulating 
powers  fully  as  much  as  the  nature  of  these  powers  seemed  to  permit. 
But  what  precisely  were  the  things  that  got  out  of  hand  ?  The  immediate 
trouble  was  with  the  stock  market  and  with  loans  on  account  of  others. 
We  need  only  follow  up  this  clue  in  order  to  get  an  answer  which  will 


1919-1929  903 

supply  also  the  qualifications  to  be  added  to  the  above  statements  about 
reserve  policy  and  its  effects.  The  loans  on  account  of  others  were  but 
a  symptom  of  the  abundance  of  money  that  prevailed  throughout  and  of 
the  powers  of  the  deposit  manufacturing  engine.  The  increase  in  loans 
and  investment  of  All  Banks  which  occurred  from  1922  to  1929  amounted 
to  about  (difference  of  annual  figures)  18.5  billions  and  tells  but  part  of 
the  tale.  The  other  part  is  in  the  fact  that  at  any  given  moment 
funds  in  existence  were  much  above  business  requirements,  the  excess 
being  uncontrolled  and  uncontrollable.  So  long  as,  under  the  influence 
of  the  lesson  of  1921,  their  owners  and  the  banks  kept  discipline  it  was 
possible  for  the  reserve  system  to  maneuver  successfully.  As  soon  as 
they  ceased  to  follow  the  lead,  as  they  were  bound  to  do  sooner  or  later, 
the  consequences  followed  automatically.  We  are  not  going  to  discuss 
responsibilities  or  the  question  whether  the  diiignosis  should  be  expressed 
by  saying  that  there* was  potential  "inflation"  all  along,  which  turned  into 
actual  "inflation"  in  1928.  It  is  enough  that  this  course  of  events, 
whatever  its  appropriate  name,  could,  though  not  by  routine  methods, 
have  been  prevented  in  1922  without  anyone  except  speculators  being 
the  worse  for  it — even  at  the  time — and  that  this  necessarily  would  have 
made  a  great  difference  in  the  intensity  of  the  depression. 

d.  The  policy  of  the  Bank  of  England  and  the  conditions  in  the 
English  open  market  may  be  described  as  they  have  been  above  (Sec.  C) 
in  terms  of  the  effort  made,  first  with  increasing,  then  with  decreasing 
success,  to  maneuver  back  to  the  gold  standard  at  prewar  parity  and  to 
keep  on  it.  As  has  also  been  observed  already,  this  did  mean  pressure 
on  the  economic  process.  And  gold  movements  and  states  of  the  gold 
reserve,  which  was  no  longer  protected  by  an  unchallenged  world-banker 
position  as  it  had  been  in  prewar  days — no  longer,  especially,  by  a  thick 
wall  of  well-disciplined  and  quick  short-term  assets  in  foreign  countries — 
hence  also  money  rates,  were  more  important  and  more  delicate  to  handle 
than  ever.  The  central  bank  aspects  of  this  are  obvious  from  Chart  LX. 

That  situation,  complicated  as  it  was  by  the  practical  necessity  of 
resuming  long-term  foreign  lending,  especially  within  the  empire,1  was 
as  difficult  in  practice  as  it  was  unproblematical  in  theory.  If  the  policy 
of  the  Bank  of  England  had  not  been  given  such  undue  prominence, 
both  in  the  analysis  of  the  shorter  fluctuations  of  English  business  and 
employment  and  in  the  causation  of  the  great  depression,2  no  further 
comment  would  be  necessary.  As  it  is,  we  must  stay  for  a  moment. 

The  Bank  of  England  encountered,  within  its  general  policy,  which 
was  dominated  by  the  exigencies  of  that  maneuvering,  current  business 
situations  in  exactly  the  same  spirit  and  in  much  the  same  way  as  it 
had  before  the  war.  In  1920  it  acted  with  less  promptness  and  went 
less  far  than  on  some  previous  occasions,  but  still  raised  its  rate  to  7  per 


904 


BUSINESS  CYCLES 


cent  on  Apr.  15  and  then  took  22  months  to  return  (Feb.  16,  1922)  to  4.5. 
However,  the  gloom  of  English  business  prospects,  which  could  not  have 
been  relieved  by  money  at  2  per  cent,  and  the  brightness  of  American 


PRO PORT  ON 
DEPOSITS  AMD  BA^K  PbST  ftlLLJ 


\ 


tOTES 


LONCON   HANK 


PHER' 


f  ROPO  TTIOh 


JALA  slCES 


B^LANCESTC  RESfRVE 


\ 


LOhDON 


\ 

BANKERS' 


NOTE 


HER" 


V 


r~\ 


SECU  ^ITIE 


OF  P  JBLIC 


\ 


1920  1925  1930 

CHART  LX. — Bank  of  England  series  (see  Appendix,  p.  1077). 

prospects,  which  could  not  have  been  darkened  by  money  at  7  per  cent, 
obviously  offer  a  much  more  plausible  explanation  of  the  difference  in  the 

1  The  relative  importance  of  loans  to  the  Dominions  or  to  borrowers  in  the  Dominions 
greatly  increased  in  the  postwar  decade,  though  total  long-term  foreign  lending  never 
came  near  the  1913  figure  (197.6  million  pounds).     Loans  to  borrowers  in  the  British 
possessions  amounted  (in  million  pounds)  to  69.2  in  1925,  57.5  in  1926,  98.2  in  1927,  81.5  in 
1928.     The  balance  of  payments  for  the  same  years  was:  +54,  +9,  -j-114,  +149. 

2  The  chief  authority  to  sponsor  what  seems  to  the  writer  a  seriously  wrong  view  is 
Mr.  Hawtrey.     C/.,  for  example,  Trade  Depression  and  the  Way  Out,  new  ed.,  1938,  pp. 
20  et  «<?<7.,  where  he  presents  a  sketch  of  English  postwar  business  fluctuations  viewed 
almost  entirely  as  functions  of  bank  rate,  which  in  turn  is  made  a  function  of  the  absorption 
of  gold  by  France  and  the  United  States.     Since  the  bank,  considering  the  precarious 
position  of  the  pound,  had  to  be  very  cautious  in  following  its  traditional  policy  of  lowering 
the  rate  in  negative  phases  and  to  keep  it  low  in  recovery,  and  very  prompt  in  raising  it  in 
prosperity,  it  is  not  difficult  to  establish  correlation  or  lagged  correlation.     But  we  know 
that  this  proves  absolutely  nothing  for  the  presence  of  a  cause-effect  relation. 


1919-1929  905 

speed  of  recovery  in  the  two  countries.  After  that,  cheap  money  policy 
prevailed  for  a  time,  as  it  had  always  done  after  the  passing  of  a  "deep" 
depression,  though,  under  the  circumstances,  it  was  not  possible  to  carry 
it  quite  so  far  as  it  had  been  carried  in  the  comparable  segment  of  the 
preceding  Kondratieff.1  For  the  next  3  years,  i.e.,  before  the  pound 
reached  parity,  the  rate  was  on  the  average  actually  lower  than  after, 
which  is  very  natural  because  maneuvering  up  to  parity  was  a  much 
easier  task  than  maneuvering  to  keep  on  it.  The  latter  had  to  be  done, 
not  only  without  the  help  of  international  speculation,  but  in  the  face 
of  constant  danger  of  attacks  from  it — for  selling  short  was  thenceforth 
(beyond  costs)  riskless.  In  the  domestic  situation  there  was,  in  fact, 
nothing  to  prevent  the  Bank  from  going  on  with  the  rate  of  3  per 
cent,  which  had  been  reached  on  July  13, 1922,  and  retained  to  July  5, 1923, 
although  the  Federal  Reserve  System  sold  United  States  securities2 
during  that  year.  But  eventually  the  strain  proved  too  much  and  on 
July  5,  1923,  the  Bank's  rate  went  to  4  per  cent,  although  the  Federal 
Reserve  System  then  discontinued  sales.  It  remained  there  until  Mar.  5, 
1925,  in  spite  of  the  Federal  Reserve  System's  buying  operations  in 
1924 — presumably,  in  order  to  enhance  their  effect  on  the  Bank's  posi- 
tion— and  also  in  spite  of  the  sales  that  followed  them.  But  then  came, 
with  only  a  2  months'  interruption,  over  2  years  at  5  per  cent.  This  rate 
was  no  doubt  higher  than  would  have  prevailed  without  the  abnormalities 
in  the  situation.  But  recovery  turned  into  prosperity  all  right,  and  if  this 
term  be  objected  to,  the  behavior  of  the  index  of  production  is,  in  any 
case,  conclusive  evidence  of  a  considerable  improvement,  which,  it  is 
true,  was  seriously  impaired  by  the  great  social  struggle  of  1926.  In 
1927  the  easy  money  policy  in  New  York  was  cautiously  responded  to 
by  a  reduction  of  bank  rate  to  4.5  per  cent,  which,  moreover,  was  not 
always  made  effective.  The  year  1929,  of  course,  revealed  the  intrinsic 
weakness  of  the  situation.  The  increase  of  the  bank  rate  to  6.5  per  cent, 
on  the  surface  not  more  serious  than  its  apparently  similarly  conditioned 
peak  in  1899,  and  the  decline  of  the  Bank's  gold  stock  below  the  Cunliffe 
level  ushered  in  the  death  struggle  of  the  pound.  But  it  no  more  caused 
the  depression  than  a  similar  rate  had  caused  a  depression  in  1899. 

1  If  1865  to  1874  be  accepted  as  a  roughly  comparable  segment,  we  may  note  that  the 
monthly  fluctuations  of  the  average  for  those  10  years  were  between  a  maximum  of  4  pounds, 
8  shillings,  8  pence  (November)  and  a  minimum  of  3  pounds,  2  shillings,  2  pence  (Septem- 
ber).    For  1875  to  1884  the  monthly  maximum  of  the  10-year  average  was  between   3 
pounds,  18  shillings,  2  pence  (November)  and  2  pounds,  15  shillings,  8  pence  (April).     See 
Palgrave,  Bank  Rate  and  Money  Market,  p.  97. 

2  For  our  theory  is :  when  the  Federal  Reserve  System  sold,  this  stiffened,  other  things 
being  equal,  the  New  York  open  market  and  attracted  gold  to  America,  which  would  also 
tend  to  stiffen  the  London  market  and,  moreover,  weaken  the  pound  and  vice  versa  in 
the  case  of  buying.     In  fact,  European  influence  was,  obviously  on  the  same   theory, 
exerted  in  favor  of  buying  operations  in  1924  and  1927.     To  some  students  this  will  seem 
too  evident  to  be  worth  stating,  to  others  it  will  seem  wrong. 


CHAPTER  XV 

The  World  Crisis  and  After1 


A.  The  World  Crisis  and  the  Cyclical  Schema. — We  should  now  be 
able  to  answer  the  question  how  far  the  course  of  events  from  the  fall  of 
1929  to  the  summer  of  1938  can  be  described  in  terms  of  the  analytic 
model  presented  in  this  book  and  how  far  other  factors,  external  or  inter- 
nal, new  or  old,  must  be  relied  on  for  explanation.  From  the  standpoint 
of  this  question  alone  and  merely  as  a  last  exercise  in  application  of  that 
model  we  will  review  the  sequence  of  situations  and  some  of  the  policies 
of  that  period.  This  object  is  not  so  modest  as  it  might  seem.  The 
answer  will,  directly  and  by  implication,  economically  and  sociologically, 
cover  much  more  ground  than  the  question  at  first  sight  suggests.  But 
precisely  because  it  will,  the  reader  should  be  reminded  once  more  that 
no  conclusion  we  may  arrive  at  can  do  more  for  him  than  to  help  him  to 
see  things  more  clearly.  From  that  to  practical  evaluations  or  recom- 
mendations the  way  is  long  indeed.  And  everyone  of  us  must,  carrying 
his  own  individual  load  of  desires,  prejudices,  and  visions,  travel  it  alone. 

Whenever  any  set  of  propositions  and  observations  leads  us  to  expect 
a  certain  event,  the  actual  occurrence  of  that  event  will  always  strengthen 
our  confidence  in  those  propositions.  In  such  cases  we  are  in  the  habit  of 
saying — at  the  risk,  as  we  have  had  ample  opportunity  of  observing,  of 
some  violence  to  logic — that  they  are  "verified"  by  that  event  or  that 
they  "explain"  it.  Now  in  this  sense  the  occurrence  at  that  time  of  a 
severe  and  prolonged  depression  in  itself  verifies  or  ratifies  the  application 
of  our  model,  i.e.,  warrants  explanation  in  terms  of  our  process.  For  we 
need  only  survey,  in  the  light  of  our  interpretation,  the  developments 
since  1898  in  order  to  understand  why  such  a  depression  should  have 
occurred  as  part  and  parcel  of  that  process.  No  claims,  be  it  repeated, 

1  Although  an  endeavor  will  be  made  at  the  expense  of  some  repetition  to  protect  the 
following  exposition  from  certain  misunderstandings  to  which  some  statements  to  be  made 
are  liable  if  taken  by  themselves,  this  chapter  should  not  be  read  before  the  analytical 
apparatus  presented  in  this  book  has  been  fully  mastered:  the  theory  of  the  "world  crisis" 
begins  on  page  one.  No  new  charts  will  be  presented,  but  those  that  have  been  used  in 
the  preceding  chapter  cover  and  illustrate  the  main  points  of  the  argument  of  this,  and 
should  be  referred  to  again. 

906 


THE  WORLD  CRISIS  AND  AFTER  907 

are  made  for  our  three-cycle  schema  except  that  it  is  a  useful  descriptive 
or  illustrative  device.  Using  it,  however,  in  that  capacity,  we  in  fact  got 
(in  Chap.  XIV,  Sec.  E),  ex  visu  of  1929,  a  "forecast"  of  a  serious  depres- 
sion embodied  in  the  formula:  coincidence  of  depression  phases  of  all  three 
cycles.  It  will  be  well  to  recall  once  more  what  this  formula  precisely 
means.  For  reasons  we  know,  capitalist  evolution  spells  disturbance. 
We  also  know  that  it  spells  simultaneous  disturbances  of  different  order  of 
importance  and  different  range  in  time.  Junctures  therefore  occur  in 
which  the  symptoms  incident  to  scrapping  and  rearranging  dominate  the 
scene.  Among  these  junctures  there  are  some  in  which  adjustments  to 
long-range  and  more  fundamental,  and  adjustments  to  short-range 
and  less  fundamental  industrial  changes  do  not  occur  at  the  same  time, 
and  there  are  others  in  which  they  do.  In  the  first  case,  symptoms 
will  be  mitigated;  in  the  second,  intensified — or,  to  return  to  the  schema, 
in  the  first  case  the  depressive  phases  of  one  or  two  cycles  will  hit  a 
"floor"  provided  by  the  nondepressive  phases  of  the  other  or  the  two 
others;  in  the  second  case  there  is  no  such  floor  and  hence  a  more  serious 
and  especially  broader  trough1 — all  of  which  could  easily  be  translated 
into  terms  of  hard  business  fact. 

In  XIV,  Ewe  went  a  step  further  by  making  the — absurd — assumption 
of  strict  periodicity  of  all  the  cycles  and  equal  duration  of  all  their  phases. 
Counting  on  in  the  same  manner,  we  should  get  a  Juglar  depression — on 
the  Kondratieff  depression  which  dates  from  the  fall  of  1925 — from  July 
1930  to  the  middle  of  November  1932,  which  should  be  severe,  to  be 
followed  by  a  Juglar  recovery  to  March  1935,  which  we  should  expect  to 
be,  owing  to  its  position  within  the  Kondratieff,  slow  and  weak.  That 
depression  would  contain  three  Kitchin  phases,  a  depression  to  the  middle 
of  April  1931,  a  revival  to  January  1932,  and  a  prosperity  to  the  middle  of 
November  1932;  and  the  latter  two  should  assert  themselves  mainly 
(not  wholly)  by  decrease  in  rates  of  decrease.  This  schema  has  none 
except  illustrative  significance  and  no  value  attaches  to  the  dates.2 

1  That  the  situation  of  1980  answers  well  to  a  description  in  terms  of  the  three-cycle 
schema  has  been  noticed  by  Professor  Alvin  Hansen,  Economic  Stabilization  in  an  Unbal- 
anced World,  1932,  p.  95. 

2  But  we  may  still  compare  at  least  the  general  import  of  our  schema  with  the  predic- 
tions made  by  forecasters  in  1930.     Professor  Warren  Persons,  for  instance,  (Forecasting 
Business  Cycles,  p.  44)  predicted  the  upturn  for  February  to  April  1931.     This  was  by 
no  means  so  compromising  a  failure  as  people — and  Professor  Persons  himself — thought. 
It  was  arrived  at  in  a  perfectly  workmanlike  way  by  methods  which  it  was  quite  possible 
to  defend.     It  is  a  not  uninteresting  question  to  ask  what  the  failure  was  due  to.     We 
observe  that  Professor  Persons'  dating  corresponds  fairly  well  with  ours  as  to  the  Kitchin 
revival.     This  is  more  than  a  coincidence.     He  was  reasoning  on  the  course  of  short 
cycles — though  not  exactly  of  our  Kitchin — which  was  all  he  recognized.     For  this  his 
forecast  was  not  wrong.     What  he  overlooked — as  businessmen  did — was  the  drift  of 


908  BUSINESS  CYCLES 

It  is,  however,  important  to  stress  the  common  sense  of  the  broad  diag- 
nosis thus  invested  with  a  spurious  precision  as  to  details.  Realizing 
from  historical  observation  the  extent  of  the  revolution  that  had  occurred 
in  the  industrial  structure  and  was  in  the  act  of  upsetting  its  system  of 
values,  shall  we  be  surprised  at  the  emergence  of  a  situation  in  which 
perhaps  three-quarters  of  all  businesses  in  the  United  States  (including 
farms)  had  to  face  the  necessity  of  an  adaptation  that  threatened  them 
with  economic  death?  And  is  there  really  much  to  object  to  in  the  state- 
ment that  this  situation  was  the  fundamental  fact  about  tfre  world 
crisis,  compared  with  which  all  other  factors,  however  important,  were 
after  all  but  mitigating  or  accentuating  accessories  ? 

Before  proceeding  to  qualify  and  to  elaborate,  we  will  advert  to  a 
consequence  which  follows  if  that  diagnosis  can  be  established.  Capital- 
ism and  its  civilization  may  be  decaying,  shading  off  into  something  else, 
or  tottering  toward  a  violent  death.  The  writer  personally  thinks  they 
are.  But  the  world  crisis  does  not  prove  it  and  has,  in  fact,  nothing  to 
do  with  it.  It  was  not  a  symptom  of  a  weakening  or  a  failure  of  the  sys- 
tem. If  anything,  it  was  a  proof  of  the  vigor  of  capitalist  evolution  to 
which  it  was — substantially — the  temporary  reaction.  And  in  any  case 
it  was — again,  substantially — no  novel  occurrence,  no  unprecedented 
catastrophe  expressive  of  the  emergence  of  new  factors,  but  only  a 
recurrence  of  what  at  similar  junctures  had  occurred  before. 

The  first  qualification  is  that,  so  far,  the  above  argument  covers  only 
the  course  of  events  down  to  the  bottom  of  the  depression,  which,  as  we 
shall  see,  occurred  in  the  second  half  of  1932.  Subsequent  events  raise 
problems  concerning  recovery  policies,  which  bar  any  statement  at  this 
point. 

Second,  it  should  be  recalled  that  we  never  either  undertook  to 
explain  or  succeeded  in  explaining  everything  about  any  crisis  or  even 
depression.  There  was,  in  particular,  the  important  class  of  "under- 
standable but  nonessential  incidents."  In  this  instance  they  may  be 
exemplified  by  the  activities  of,  say,  Hatry  or  Kreuger  and  so  on:  even 
German  experience  might  have  been  somewhat  different  but  for  the 
impulsive  personality  of  the  leading  man  of  the  Darmstadter  und  Nation- 
albank.  An  example  of  more  important  elements  of  this  class  of  phe- 
nomena is  the  violence — though  not  simply  the  occurrence — of  the  boom 
and  the  crash  in  the  stock  market  of  the  United  States,  which  forms  no 
part  of  the  essentials  of  our  process,  yet  powerfully  influenced  it.  But 
the  boundaries  of  this  class  should  not  be  extended  too  far.  The  building 

things  below  that  surface  movement,  i.e.,  the  longer  cycles  and  their  phases.  That  failure 
of  predictions  to  come  true  may  thus  be  used  to  illustrate  the  shortcomings  of  what  we  have 
termed  the  single-cycle  hypothesis.  It  also  illustrates  the  necessity  for  forecasting  of 
understanding  the  logic  of  industrial  evolution. 


THE  WORLD  CRISIS  AND  AFTER  909 

booms  and  their  slackening  from  about  1928  on,  for  instance,  or  the  bulk 
of  the  difficulties  in  the  agrarian  sectors  do  not  belong  here  but  were 
perfectly  normal  elements  of  that  process. 

The  American  debt  situation  and  the  American  bank  epidemics — 
there  were  three  of  them — are  in  a  class  by  themselves.  Given  the  way 
in  which  both  firms  and  households  had  run  into  debt  during  the  twenties, 
it  is  clear  that  the  accumulated  load — in  many  cases,  though  not  in  all, 
very  sensitive  to  a  fall  in  price  level — was  instrumental  in  precipitating 
depression.  In  particular,  it  set  into  motion  a  vicious  spiral  within 
which  everybody's  efforts  to  reduce  that  load  for  a  time  only  availed  to 
increase  it.  There  is  thus  no  objection  to  the  debt-deflation  theory  of  the 
American  crisis,  provided  it  does  not  mean  more  than  this.1  The  element 
it  stresses  is  part  of  the  mechanism  of  any  serious  depression.  But 
increase  of  total  indebtedness  at  the  rate  at  which  it  had  occurred  in  this 
country  is  neither  a  normal  element  of  the  mechanism  of  Kondratieff 
downgrades — repayment  such  as  was  effected  by  big  concerns  either  from 
profits  or  from  the  proceeds  of  bond  and  stock  sales  fits  better  into  the 
picture — nor  in  itself  an  "understandable"  incident,  like  speculative 
excesses  and  the  debts  induced  by  these.  It  must  be  attributed  to  the 
humor  of  the  times,  to  cheap  money  policies,  and  to  the  practices  of 
concerns  eager  to  push  their  sales;  and  it  enters  the  class  of  understand- 
able incidents  only  if  we  include  specifically  American  conditions  among 
our  data.  Similarly,  bank  failures  are  of  course  very  regular  (though  still 
not  essential)  occurrences  in  the  course  of  any  major  crisis  and  invariably 
an  important  cause  of  secondary  phenomena,  in  particular,  again,  of 
downward  cumulative  processes.  Those  epidemics  cannot,  however, 
any  more  than  the  German  difficulties  in  this  field,  be  considered  as 
wholly  explained  by  the  ordinary  mechanism  of  crises  or  by  that  mecha- 
nism plus  the  fact  of  excessive  indebtedness  all  round  or  even  by  all  that 
plus  the  stock  exchange  crash.  The  German  breakdown  is  partly 
attributable  to  very  special  circumstances  of  an  extra-economic  nature 
without  which  nothing  could  have  shaken  such  concerns  as  the  Austrian 
Kreditanstalt  or  the  Dresdner  Bank.  The  American  epidemics  become 
fully  understandable  only  if  account  be  taken  of  the  weaknesses  peculiar 
to  the  American  banking  structure,  which  made  it  succumb  as  no  Euro- 
pean system  would  have  succumbed  under  similar  circumstances — in 
particular,  the  presence,  fostered  by  legislation  and  public  opinion,  of  a 
large  number  of  small  and  inefficient  banks  and  the  absence  of  anything 
like  the  English  tradition.  Compared  with  this,  the  insolvency  of  foreign 

1  Professor  Irving  Fisher,  see,  for  example,  Booms  and  Depressions,  p.  85,  does  not 
seem  to  claim  substantially  more  than  that.  We  may  demur  at  his  sweeping  statement 
that  a  fall  in  prices  always  impairs  debtors'  ability  to  pay.  But  "that  over-indebtedness 
and  deflation  were  strong  and  indeed  dominating  factors"  need  not  be  denied. 


910  BUSINESS  CYCLES 

debtors  was,  especially  since  it  impinged  primarily  on  strong  concerns,  of 
minor  importance. 

Third,  there  were  the  external  factors.  In  noticing  them  in  Chap. 
XIV,  Sec.  C,  we  found  that  their  importance,  so  far  as  the  causation 
of  the  world  crisis  is  concerned,  may  easily  be  overestimated.  Some  of  the 
most  conspicuous  political  and  social  changes  of  the  postwar  time,  such 
as  the  Russian  developments  or  the  destruction  of  the  Austro-Hungarian 
Monarchy,  no  doubt  interfered  with  the  "normal"  working  of  the 
economic  engine  but  have — excepting  perhaps  the  breakdown  of  the 
Viennese  banking  center — little  if  any  bearing  on  the  crisis.  Shifts  of, 
and  impediments  to,  international  trade  made  many  countries,  and 
especially  England,  less  prosperous  than  they  would  otherwise  have 
been,  and  shaped,  by  thus  weakening  the  organisms  on  which  the  crisis 
impinged,  very  many  details  of  the  picture,  but  this  is  all.  In  particular, 
it  would  be  unwarranted  to  attribute  any  of  the  major  features,  let  alone 
causation,  of  the  American  depression  to  a  "flood  of  imports."  Imports, 
on  the  contrary,  fell  rapidly  at  the  critical  time:  they  totaled  $4.4  billions 
in  1929,  little  over  3  in  1930,  a  little  over  2  in  1931,  and  about  1.3  in  1932. 
German  exports  were  in  1929  and  1930  a  nonnegligible  factor.  But  they 
and  their  possible  effects  were  more  a  consequence  than  a  cause.  Nor 
can  the  fiscal  and  social  policies  of  England  and  Germany  have  directly 
accounted  for  more  than  a  weakened  resistance  to  the  impact  of  depres- 
sion1— against  which,  moreover,  must  be  set  their  restraining  influence 
during  preceding  prosperities. 

We  have  seen,  it  is  true,  that  the  indirect  influence  of  those  policies, 
certainly  in  the  case  of  Germany  and  possibly  also  in  the  case  of  England, 
on  short  capital  movements  had  much  to  do  especially  with  what  in  a 
narrower  and  perhaps  more  proper  sense  should  be  called  the  crisis. 
And  the  provisional  solution  that  had  been  arrived  at  for  the  problem  of 
international  payments  was  bound  to  break  down  in  any  major  depression 
and,  before  doing  so,  to  accentuate  its  difficulties.  We  do  not  even 
wish  to  palliate  the  influence  of  such  monetary  disorders  as  occurred 
before  the  depression  had  the  chance  of  tearing  up  the  flimsy  tissue 
of  pseudo  gold  currencies.  The  South  American,  especially  the  Argen- 
tinian, disorders  and  the  fall  of  the  price  of  silver  have  undoubtedly 
played  some  role.2  But  all  this  looms  so  large  because  a  depression 

1  This  is  perfectly  compatible  with  our  estimate  of  their  importance  in  general.     The 
above  statement  is  intended  only  to  guard  against  overestimation  of  their  importance 
in  the  causation  of  the  great  depression. 

2  Gold  movements  are  covered  by  short  capital  movements.     In  spite  of  devaluation, 
the  shrinkage  of  international  trade,  and  repayments,  the  Bank  for  International  Payments 
estimated  the  total  of  international  short  credits  at  from  29  to  30  billion  Swiss  francs  for 
1934,  roughly  three  times  the  prewar  figure.     But,  as  has  been  stated  before,  apart  from 
them  gold  cannot  have  played  any  major  role. 


THE  WORLD  CRISIS  AND  AFTER  911 

occurred  for  other  reasons.  As  a  man  may  suffer  from  many  ills  and  yet 
for  an  indefinite  time  lead  a  vigorous  life  without  being  seriously  incon- 
venienced by  them  until,  when  his  general  vitality  has  ebbed  away,  those 
ills  or  any  one  of  them  may  suddenly  acquire  what  to  the  specialists'  eye 
will  seem  paramount  or  even  fatal  importance;  so  the  economic  organism 
always  does  bleed  from  many  wounds  which  it  bears  lightly  in  three  out 
of  the  four  cyclical  phases,  and  which  spell  discomfort  when  one  cycle, 
distress  when  two,  catastrophe  when  all  the  cycles  are  in  the  depression 
phase.  No  doubt,  external  injuries  were  of  unusually  great  importance  in 
this  case,1  but  explanation  cannot  be  derived  from  them. 

B.  1930.  1.  The  United  States. — The  businessmen  and  forecasters, 
who  in  the  fall  of  1929  had  made  up  their  minds  that  nothing  worse  was 
ahead  than  a  "recession"  not  much  more  serious  than  that  of,  say,  1924, 
cannot  have  been  very  disappointed  by  the  general  look  of  things  during 
the  first  half  of  1930.  At  the  beginning,  stock  prices  rallied  strongly, 
security  issues  were  large,  signs  of  improvement  showed  in  many  spots, 
money  was  easy.  All  that,  except  the  easy  money — prime  commercial 
paper  rate  fell  from  4.89  per  cent  in  January  to  £  per  cent  in  August  and 
was  2.88  in  December;  for  June  it  was  3.54  per  cent — passed  quickly, 
it  is  true,  and  proved  to  have  been  a  meaningless  flurry,  due  perhaps  to 
the  confidence  that  was  so  widely  felt  or  to  the  organized  effort  to  make  a 
stand  that  prompted  some  additional  spending.  But  until  about  the  end 
of  June  business  moved  along  on  a  but  slowly  falling  level  not  much 
below  the  figures  of  1929  in  practically  all  lines.2 

The  second  half  of  the  year  presents  a  wholly  different  picture.  What 
was  generally  recognized  as  liquidation  all  round  was  the  outstanding 
feature.  Rates  of  contraction  quickened  and  comparison  with  1929 
figures  became  increasingly  melancholy.  As  we  have  put  it  above 
(Chap.  XIV,  last  sentence  of  Sec.  E)  people  felt  that  the  ground  under 
their  feet  was  giving  way.  There  was,  however,  no  panic  or  even  alarm 
until,  late  in  the  year,  distress  signals  showed  in  the  banking  sphere,  the 
failure  of  the  Bank  of  the  United  States  (December)3  attracting  particular 
attention  abroad.  And  not  only  totals  for  the  year  but  also  figures  for 
the  end  of  the  year  were  far  from  being  catastrophic.  Total  corporate 
issues  were  $5.473  billions — slightly  above  the  level  of  1926,  though  only 

1  It  should  not  be  forgotten,  however,  that  the  crisis  was  nowhere  else  anything  like 
so  severe  as  in  the  United  States,  the  country  most  nearly  free  from  injury  by  external 
factors. 

2  The  business  curve  of  the  Harvard  Barometer  was,  during  those  6  months,  almost 
horizontal.     The  A  curve  ("speculation")  had  lost  by  May  what  ground  it  had  gained 
during  the  first  quarter. 

3  The  writer  has,  however,  not  been  able  to  convince  himself  that  Europeans  really 
thought  that  this  bank  was  something  like  the  Bank  of  England  or  the  Banque  de  France, 
and  that  this  belief  materially  influenced  events. 


912  BUSINESS  CYCLES 

about  55  per  cent  of  the  1929  figure — or,  exclusive  of  refunding  issues, 
4.5  billions,  slightly  below  the  level  of  1926.  Outside  debits  (133  cities) 
with  137.5  billions  were  only  a  little  below  1926  and  about  14  per  cent 
below  1929,  though  contraction  was  severe  from  the  beginning  of  July 
to  the  beginning  of  September.  Outside  net  demand  deposits,  as  already 
pointed  out  in  the  preceding  chapter,  neither  were  appreciably  lower  for 
the  year  nor  fell  appreciably  within  the  second  half  of  the  year — July, 
8.117  billions;  December,  7.911  billions.  But  All  Other  Loans  had  already 
shrunk  sharply  from  November  1929  to  May  1930  and  then  continued  to 
decline.  Some  hoarding  demand  for  money,  increasing  "money  in  circu- 
lation" and  total  federal  reserve  credit  outstanding,  showed  in  December. 
Number  of  banks  suspended  (1,345)  was  in  fact  more  than  double  the 
yearly  average  1921-1929  (627),  but  1,158  of  them  were  nonmembers.1 
Excepting  July  and  August,  there  was  net  influx  of  gold  in  every  month 
which,  together  with  the  issue  of  additional  national  bank  notes,  brought 
excess  reserves  of  member  banks  to  about  $475  millions  by  the  middle  of 
November — a  position  of  the  banking  system  as  a  whole  that  was  tech- 
nically anything  but  weak,  although  the  value  of  collateral  was  already 
seriously  impaired. 

There  was  nothing  abnormal  under  the  circumstances  in  the  quicken- 
ing decline  of  prices  of  finished  products,  which  within  the  year  fell  by 
about  10  per  cent.  The  wholesale  price  index  was  pulled  down  by  the 
fall  in  the  prices  of  semifinished  goods  and  especially  of  raw  materials — • 
the  December  average  of  the  latter  was  over  20  per  cent  below  the  Jan- 
uary average.  Money  rate  of  wages  fell  considerably  in  agriculture,  but 
was  (see  below)  substantially  maintained  in  industry.  In  the  fourth 
quarter,  however,  reductions  in  individual  industries  were  sufficiently 
important  to  affect  the  general  index,  though  they  still  left  real  hourly 
earnings  at  a  level  higher  than  that  of  1929.  Weekly  money  earnings 
decreased  even  for  the  first  quarter,  and  weekly  real  earnings  eventually 
fell2  to  90  per  cent  above  the  prewar  level.  Wage  bill  continued  its 
descent  from  the  peak  of  the  third  quarter  of  1929.  In  New  York  State, 
for  instance,  factory  pay  rolls  declined  by  about  25  per  cent  from  January 
to  December. 

Total  monetary  labor  income  for  the  year  as  estimated  in  the  study 
made  by  the  Department  of  Commerce  in  cooperation  with  the  National 
Bureau  (S.  Kuznets,  National  Bureau  Bulletin,  Jan.  26,  1934,  p.  5)  was, 
however,  only  7.9  per  cent  and  total  monetary  income  produced  only 
15.1  per  cent  below  1929.  But  net  earnings  of  all  corporations  (excluding 

1  The  sum  of  deposits  affected  was,  however,  nearly  865  millions,  more  than  four  times 
the  1921  to  1929  average. 

2  For  the  cost  of  living  index  used  in  arriving  at  the  above  statements  see  International 
Abstract  of  Economic  Statistics,  1934,  p.  205. 


THE  WORLD  CRISIS  AND  AFTER  913 

tax-exempt  ones  and  life  insurance  companies)  before  deduction  of 
income  taxes  were  by  over  78  per  cent  below  the  1929  figure  ($1,960 
millions),  printing  and  publishing,  foods,  beverages,  tobacco  products, 
chemicals,  metal  and  metal  products,  paper,  pulp  and  its  products  doing 
comparatively  well,  textiles  particularly  badly.1  The  most  serious 
features  of  the  picture  are  displayed  by  the  indices  of  industrial  produc- 
tion and  of  employment.  Output  of  equipment  and  durable  goods  in 
general,  as  reflected,  for  example,  in  steel  ingot  production,  fell  sharply 
after  May.  Motor- vehicle  factory  sales  were  less  than  2.8,  as  against 
4.6  billions.  The  Harvard  Society's  index  of  volume  of  manufactures 
declined  by  over  22,  the  Federal  Reserve  Board's  index  of  employment 
by  over  16.2  per  cent  within  the  year,2  whereby  the  former  almost 
reached  the  trough  figures  of  192 13  and  the  latter  slightly  fell  below  them. 

Mining  did  somewhat  better  than  manufactures,  and  power  produc- 
tion was  in  the  first  six  months  above,  and  in  the  last  six  months  not  much 
below,  the  1929  level.  Total  construction  as  measured  by  contracts 
(Dodge  figures  for  37  states)  was  over  20  per  cent  below  1929  and  would 
have  been  still  lower  if  publicly  and  semipublicly  financed  building  had 
not  kept  up  or  even  somewhat  increased.  But  it  is  worth  noting  that 
public  utilities  increased  their  expenditure  on  construction  ($644  millions 
as  against  473  in  1929)  and  that  residential  building  did  not  decline  much. 
It  was,  as  we  should  expect,  industrial  and  commercial  building  which 
caused  the  fall  in  the  total. 

The  brightest  features  of  the  picture  are  to  be  found  in  the  sphere  of 
consumption,  the  surprisingly  good  showing  of  which  has  not  always 
received  the  attention  it  merits.  Sales  by  department  stores,  while  con- 
sistently and  increasingly  below  those  of  1929,  were  still  102  per  cent  of 
the  1923  to  1925  average  and  their  December  holiday  trade  in  the  New 
York  district — containing,  however,  one  more  selling  day  than  in  1929 — 
was  only  4.5  per  cent  under  that  of  the  previous  year.  Considering  the 
fall  in  prices  there  obviously  cannot  have  been  decline  in  physical  volume. 
This  is  borne  out  by  the  behavior  of  carloadings  of  the  l.c.l.  category, 
which  in  contrast  to  total  carloadings  declined  but  moderately.  Con- 
sumption of  a  number  of  articles,  such  as  cigarettes  and  gasoline,  electric 

1  See  S.  Fabricant,  National  Bureau  of  Economic  Research  Bulletin,  Apr.  11,  1985. 

2  16.2  is  the  percentage  by  which  the  December  figure  was  below  the  January  figure, 
which  is  what  we  mean  by  loss  within  the  year.     The  production  index,  however,  did  not 
decline  monotonically,  and  the  January  figure,  though  the  index  is  corrected  for  seasonal 
variation,  does  not  seem  to  represent  the  situation  well,  because  it  reflects  restrictions 
decided  on  at  the  end  of  1929  and  partly  reversed  at  the  very  time  when  they  affected  the 
index.     So  the  average  of  the  first  3  months  was  taken  instead. 

3  Wheat,  cotton,  cattle  and  hog  receipts  were  but  little  below  1929:  the  agrarian  sector 
behaved  according  to  form.     The  value  of  the  wheat  crop  was  only  a  little  over  60  per  cent, 
that  of  the  cotton  crop  less  than  60  per  cent,  of  the  respective  values  in  1929. 


914  BUSINESS  CYCLES 

current  for  domestic  use,  telephones,  radio  sets,  and  refrigerators, 
increased  or  declined  but  insignificantly.  To  save  space,  reference  is 
made  to  the  careful  study  of  Mr.  A.  B.  Tebbutt.1  The  number  of 
business  failures,  consistently  and  increasingly  above  that  of  1929,  was 
as  yet  far  from  alarming:  about  2,000  a  month,  which  is  less  than  the 
average  number  of  failures  from  October  1921  to  June  1922. 

An  attempt  to  answer  the  question  how  far  these  facts  can  be  trusted 
to  reflect  the  working  of  our  model  (including  "understandable  non- 
essentials")  naturally  divides  up  into  two  tasks.  On  the  orte  hand,  we 
have  to  ask  whether  expectations  from  our  model  are  adequately  borne 
out.  This  is  obviously  the  case.  Even  disregarding  the  exact  coinci- 
dence of  dates  with  our  experimental  count,  we  readily  see  not  only  that 
the  history  of  that  year  as  a  whole  is  not  badly  rendered  by  the  formula 
"a  recession  sliding  off  into  deep  depression,"  but  also  that  the  economic 
physiognomies  of  the  two  halves  of  the  year,  differing  as  they  do  char- 
acteristically from  each  other,  conform  to  our  idea  of  such  a  situation  in 
every  single  symptom  excepting  the  short-lived  rally  at  the  beginning. 
The  reader  should  have  no  difficulty  in  verifying  this  proposition  as  to 
general  contours — the  behavior  of  price  level,  output,  interest  rates, 
deposits,  clearings,  incomes,  employment,  and  so  on — and  it  will  be 
sufficient  to  draw  attention  to  a  few  points. 

That  rally  at  the  beginning  was,  as  has  been  said  and  as  will  presently 
be  explained,  due  to  organized  effort,  but  too  much  importance  should 
not  be  attributed  to  that  effort,  because  expectation  would  in  any  case 
have  been  for  fairly  sustained  business  at  that  time.  In  particular,  it  is 
perfectly  in  keeping  with  our  schema  that  industrial  money  wages  did 
not  during  the  first  half  of  the  year  fall  to  any  significant  extent — the 
reader  will  remember  that  there  is  no  reason  to  expect  wages  to  fall  in 
recession  and  that  no  inference  about  subsequent  disturbance  can  be 
drawn  from  their  failure  to  do  so.  The  ease  in  the  money  market  was 
also  normal  under  the  circumstances2  and  a  simple  consequence  of  "busi- 
ness deflating  itself."  It  does  not  call  for  explanation  by  Federal 
Reserve  Board  policy,  although  the  latter  no  doubt  contributed  to  it 
(see  below).  Not  less  true  to  form  were  the  effects  of  that  ease:  as  nor- 
mally happens  in  recession,  it  helped  to  keep  up  residential  construction 

1  Arthur  R.  Tebbutt,  Behavior  of  Consumption  in  Business  Depression,  published  as 
No.  8  of  the  Harvard  Business  School  Business  Research  Studies.  For  that  and  the  fol- 
lowing year  this  study  gives  a  very  instructive  picture,  although  its  details  (such  as  the 
behavior  of  candy  and  women's  night  attire)  are  sometimes  not  easy  to  understand.  The 
opportunity  may  be  taken  to  refer  also  to  other  studies  in  the  same  series  which,  as  for 
instance  those  on  the  operating  results  of  chain  stores,  present  valuable  information  about 
the  later  phases  of  the  depression  and  about  the  subsequent  recovery. 

a  We  observe  similar  ease  and  a  sharp  fall  in  rates  after  the  crash  in  1873  and  after  that 
of  1893. 


THE  WORLD  CRISIS  AND  AFTER  915 

and  to  induce  certain  types  of  investment,  for  example,  by  utilities,  which, 
however,  acted  also  under  another  stimulus.  The  unequal — in  the  case 
of  some,  especially  of  new,  commodities  negative — rates  of  restriction, 
the  severe  slump  in  the  second  half  of  the  year  in  the  sales  of  "post- 
ponables,"  and  the  behavior  of  consumption  should  be  particularly 
noticed. 

Hence,  although  our  methods  do  not  enable  us  to  formulate  our 
expectations  numerically  and  although  it  is  impossible  to  say  whether 
our  series  exactly  behaved  as  they  should  have  done  according  to  our 
theory,  it  is  possible  to  say  that  they  did  so  as  far  as  we  can  make  out. 
Facts  would  not  refute  us  even  if  we  made  bold  to  say  that  nothing  but 
our  process  had  acted  on  the  economic  system,  and  they  certainly  bear 
us  out  if  we  conclude  that  that  process  constituted  the  dominant  factor, 
while — witness  all  those  undoubtedly  very  competent  judges  of  business 
situations  who  were  unable  to  account  for  or  to  predict  anything  but  a 
brief  recession — without  that  process  the  events  of  that  year  and  of  its 
sequel  would  have  to  be  explained  in  terms  of  such  slogans  as  overproduc- 
tion, oversaving,  overinvestment,  and  so  on,  i.e.,  could  not  be  explained 
at  all. 

On  the  other  hand,  we  must  consider  the  possibility  that  other  factors 
influenced  the  course  of  things  in  such  a  way  as  to  produce  a  spurious 
verification.  Attempts  to  influence  the  process  have  not  been  entirely 
absent  and  must  be  taken  into  account  even  if,  relying  on  previous  argu- 
ment, we  discard  the  possibility  that  influences  from  abroad  might  have 
significantly  affected  the  American  situation  in  other  ways  than  through 
the  stock  exchange.1  There  were,  first,  measures  in  aid  of  agriculture 
and  an  appeal  to  the  household  remedy  of  the  party  in  power,  the  Hawley- 
Smoot  tariff.  The  effects  of  the  former,  though  certainly  not  negligible, 
failed  even  in  the  agrarian  sector  to  modify  conditions  sufficiently  to 
make  them  diverge  from  expectation  and  cannot  a  fortiori  have  been 
very  important  for  the  system  as  a  whole.  The  effects  of  the  latter, 
partly  counteracted  by  reprisals,  may  under  the  circumstances  be  equated 
to  zero.2 

Second,  there  was  the  President's  hortatory  action,  which  was,  by 
staving  off  reductions  in  wages  and  by  stimulating  investment,  "to  make 

1  That  influence  was  exerted  through  stock  exchange  operations  we,  of  course,  do  not 
deny.     But  this  is  sufficiently  taken  account  of  by  recognizing  that  boom  and  crash  had 
been  more  severe  than  they  otherwise  would  have  been. 

2  The  writer  does  not  wish  to  deny  that  conditions  in  certain  industries  may  have 
been  slightly  steadied  by  the  tariff  or  that  a  more  thoroughgoing  analysis  might  make  out  a 
stronger  case.     But  until  proof  to  the  contrary  is  forthcoming,  he  is  inclined  to  rate  both 
the  stabilizing  and  the  dislocating  effects  of  the  Hawley-Smoot  Act  rather  low,  and  net 
effects  still  lower. 


916  BUSINESS  CYCLES 

certain  that  the  fundamental  business  of  the  country  shall  continue  as 
usual/*1  The  interest  for  us  of  this  and  similar  attempts  consists  mainly 
in  testing  those  theories  that  see  in  depressions  nothing  else  but  the  result 
of  businessmen's  moods,  which,  themselves  ultimate  data,  shape  business 
situations  by  means  of  the  cumulations  and  accelerations  induced  by 
individual  acts.  In  this  case  conditions  were  quite  exceptionally  favor- 
able to  success.  The  American  business  world  was,  as  has  been  pointed 
out,  by  no  means  overpessimistic  at  that  time.  It  was  in  the  habit  of 
looking  for  a  lead  to  the  heads  of  a  relatively  small  number  df  concerns, 
which,  moreover,  were  big  enough  to  be  able  to  influence  the  situation 
"mechanically"  by  their  own  action.  Expansion  would  have  been 
eminently  to  their  interest  and  certainly  was  what  they  actually  wished  to 
see.  Receiving  themselves  a  lead  from  a  political  agent  that  was  in  no 
way  antipathetic  to  them  and  being,  many  of  them,  imbued  with  high- 
wage  theories,  they,  in  fact,  made  an  effort  both  by  refraining  from  lower- 
ing wages  or  from  doing  anything  else  that  would  have  been  suggestive 
of  depression,  and  by  launching  out  into  investment — the  utilities  and 
railroads  in  particular  (see  Chart  LIII)  responded  to  the  appeal  and  even 
borrowed  for  the  purpose,  so  that  the  "principle  of  acceleration"  had 
plenty  to  work  with.  Nor  was  the  result  simply  nil.  The  case  shows  to 
perfection  what  can  and  what  cannot  be  achieved — and  explained — in 
this  way.  We  have  above  noticed  the  "flurry"  at  the  beginning  of  the 
year,  which  may,  though  only  in  part,  be  attributed  to  that  effort. 

Third,  we  have  seen  that  public  expenditure  was  kept  going  and  even 
increased,  especially  by  means  of  public  construction.  It  has  been 
estimated2  that  net  federal  income-generating  expenditure  for  the  year 
was  251  and  for  the  second  half  alone  $450  millions.  This  is  not  negligible. 
But  it  may  well  be  doubted  whether  that  part  of  the  total  which  may  be 
reasonably  assumed  to  have  become  fully  effective  during  the  year  can 
have  influenced  events  materially.  The  Federal  Reserve  System,  finally, 
followed  a  policy  of  easy  money — which  policy,  as  we  have  seen,  but 
ratified  the  situation — thus  giving  business  all  the  rein  and  all  the  encour- 
agement it  could  possibly  have  wished  for.3  As  soon  as  it  had  become 

1  Cf.  President  Hoover's  message  to  Congress,  December  1929.  For  similar  utterances 
by  various  types  of  authorities  see  Mr.  Walter  Lippmann's  quotations  in  The  United  States 
in  World  Affairs,  1931. 

*  See  Professor  A.  D.  Gayer,  What  is  Ahead?,  The  New  Republic,  Feb.  2,  1938,  p.  391. 
The  federal  deficit  for  1930-1931  was  over  900  millions.  This  was  mainly — to  almost 
three-quarters — due  to  a  fall  in  revenue,  while  in  England  and  Germany  revenue  increased, 
in  England's  case  by  43  million  pounds;  in  Germany's  case  insignificantly.  Taxation  may 
by  mobilizing  idle  deposits  be  income-generating,  even  if  it  entail  no  borrowing  from  banks. 

'The  question  whether  the  "dear  money  policy"  of  1928  and  1929  can,  partly  or 
wholly,  be  held  responsible  for  the  events  in  1930  has  been  discussed  in  the  preceding 
chapter,  Sec.  F;  the  questions  of  principle  involved,  in  Chap.  XIII. 


THE  WORLD  CRISIS  AND  AFTER  917 

obvious  that  business  was  in  for  a  "recession"  and  before  any  vicious 
spiral  had  developed,  the  reserve  system  resorted  to  what  the  public  and 
many  economists  had  by  then  come  to  believe  in  as  the  remedy,  open- 
market  purchases  on  a  large  scale.  From  October  1929  to  December 
1930  it  bought  government  securities  to  the  amount  of  560  millions  or, 
more  precisely,  its  holdings  of  government  securities  rose  from  136 
millions  on  Oct.  23,  1929,  to  533  millions  on  Dec.,  18,  1929,  then  fell  to 
477  in  January  1930  to  rise  again  to  602  by  the  end  of  August.  After 
that  date,  purchases  were  reduced  to  insignificant  amounts.  For  this 
there  were  two  excellent  reasons.  First,  those  purchases  which  it  was 
thought  had  been  so  effective  in  stimulating  activity  in  1924  and  1927, 
at  this  time  did  not  seem  to  have  any  effect  at  all.1  Members  reacted 
primarily,  although  they  also  increased  their  investments,  by  paying 
off  rediscounts,  and  were  obviously  much  more  troubled  about  finding 
customers  for  their  funds  than  about  finding  funds  for  their  customers:2 
to  anyone  at  all  open  to  argument  and  evidence  on  the  subject,  any 
further  steps  in  that  direction  would  have  seemed  altogether  futile. 
Later  on,  in  the  second  place,  the  gold  influx  provided  another  reason  for 
discontinuing  that  policy.3  We  may,  hence,  conclude  that  the  behavior 
of  the  reserve  system,  while  it  favored  cheap  money  and  expansion  and 
certainly  exerted  no  depressing  effects,  yet  was  no  major  factor  in  shaping 
the  business  situation — our  process  seems,  as  far  as  that  goes,  in  fact  to 
have  worked  all  but  undisturbed. 

2.  England. — The  outstanding  fact  about  the  English  depression  is  its 
mildness,  which  makes  it  doubtful  whether  that  term  is  applicable  at  all. 
This  statement  may  sound  strange  to  many  readers  who  are,  in  connec- 
tion with  the  world  crisis,  impatient  of  anything  but  superlatives.  It 
could,  however,  be  reduced  to  the  obvious  if  we  consider  that  the  behavior 
of  some  of  the  worst  symptoms  is  amply  accounted  for  by  the  fall  in 
exports  and  the  impact  of  foreign  insolvencies — phenomena  that  were 

1  This  will  not  astonish  us.     But  it  is  not  superfluous  to  weigh  the  implications  of  that 
fact,  which  should  according  to  ordinary  rules  of  inference  suffice  to  cast  doubt  on  the  two 
previous  instances. 

2  Rediscounts  had  been  at  about  1  billion  in  the  summer  of  1929  and  were  down  to 
about  200  millions  by  July  1930.     Investments  of  reporting  member  banks  rose  from 
5,486  millions  on  Aug.  28,  1929,  to  6,829  millions  on  Aug.  27,  1930.     The  reader  should 
observe  how  very  "regular"  all  this  was  from  our  standpoint. 

8  For  the  year  net  gold  imports  were  about  $280  million.  Net  release  from  earmark  was 
minus  4.4;  total  increase  in  gold  stock  305.4  millions.  An  eminent  economist,  who  is 
more  enthusiastic  about  open-market  operations  than  the  writer  feels  able  to  be,  attributes 
their  failure  to  take  effect  to  inadequate  dosiqg  and  to  the  policy's  being  abandoned  "in 
the  face  of"  the  gold  imports  from  October  on.  This  implies  that  he  looks  upon  gold 
imports  not  only  as  factors  of  decrease  of  Federal  reserve  credit  outstanding  in  the  technical 
sense,  but  also  as  a  reason  for  and  not  against  open-market  purchases.  This  the  present 
writer  entirely  fails  to  understand. 


918  BUSINESS  CYCLES 

to  be  sure  largely  due  to  cyclical  factors  (and  their  "secular"  net  results) 
but  must  be  classed  as  external  from  the  standpoint  of  the  English 
organism.1  But,  even  independently  of  that,  the  depression  was  in 
important  respects  much  milder  than  in  this  country.  Nor  shall  we 
wonder  at  it,  for  if  the  evolutionary  process  is  the  fundamental  "cause" 
of  prosperity  and  depression,  relative  weakness  of  its  positive  phases  will 
be  accompanied  by  relative  mildness  of  negative  phases. 

In  1930  the  London  and  Cambridge  Economic  Service  annual  index 
of  total  production  fell  to  a  little  below  the  figure  of  1928  (1928,  108.7; 
1929,  115.8;  1930,  106.5).  Some  groups,  however,  displayed  increase 
(leather  and  India  rubber  trades)  or  insignificant  decrease  (food,  drink, 
tobacco;  nonferrous  metals).2  Within  the  steel  group,  the  number  of 
motor  vehicles  produced  declined  very  little.3  Analysis  of  the  building 
index,  which  fell  more  than  any  other  component  of  the  total,  reverses  the 
inference  to  be  drawn  from  it:  the  fall  in  houses  completed  during  the 
year  is  entirely  due  to  the  cessation  of  construction  under  the  Chamber- 
lain Act  and,  to  a  lesser  extent,  to  temporary  reduction  of  building  under 
the  Wheatley  Act,  while  the  joint  effect  of  this  was  to  more  than  50  per 
cent  made  up  for  by  privately  financed  construction.  This  increase  in 
unsubsidized  housebuilding,  which  gathered  force  instead  of  slackening  in 
the  second  half  of  the  year,  suffices  by  itself  to  negative  the  idea  of  a 
depression  of  unprecedented  severity.  With  this  the  high  and  rising 
unemployment  figures  seem  difficult  to  reconcile.  But  the  fall  in  exports 
(see  below)  partly  resolves  this  apparent  contradiction.4 

Wholesale  prices  and  cost  of  living  fell  steadily,  as  we  should  expect — 
the  fall  in  the  former,  the  reader  should  recall,  meant  partly,  though  not 
wholly,  a  gift  presented  to  England  by  the  countries  producing  raw  mate- 
rials— and  short  money  rates  continued  during  the  first  5  months  their 

1  It  is  disheartening  to  observe  that  difference  of  opinion  is  possible  about  what  seems 
to  the  writer  so  obvious  a  point.     Mr.  Hawtrey  (Trade  Depression  and  the  Way  Out,  p.  27) 
for  instance,  state   that  in  the  latter  part  of  1929  England  was  already  a  "deadly  centre  of 
contagion"  from  which  the  "blight  of  pessimism"  was  ready  to  spread.     This  is  on  a  par 
with  his  statement  (ibid.,  p.  26)  that  depression  had  prevailed  in  England  ever  since  1920 
(sic).     In  part  such  differences  are  presumably  due  to  different  use  of  terms.     But  the 
present  writer  feels  unable  to  understand  those  statements — and  many  others — in  what- 
ever sense  he  can  conceive  them  to  have  been  meant. 

2  The  index  includes  agricultural  output,  which  also  declined,  though  not  quite  so  much 
as  the  total  index. 

8  So  did  the  shipbuilding  index  of  the  Service.  It  is,  however,  important  to  note  that 
tonnage  commenced  declined  drastically  in  the  second  half  and  especially  toward  the 
end  of  the  year.  Coal  also  declined  (see  below),  but  less  than  steel.  Conditions  in  the 
textile  industry  were  unsatisfactory. 

4  From  January  to  December  1930  the  total  number  of  unemployed  insured  persons 
(Great  Britain  and  Northern  Ireland)  increased  by  nearly  700,000.  The  average  figure  for 
the  year  was  16.1  per  cent.  Part  of  this,  however,  was  an  effect  of  previous  rationalization. 


THE  WORLD  CRISIS  AND  AFTER  919 

descent  from  the  peak  in  the  autumn  of  1929.  After  that  they  fluctuated 
at  a  low  level,  the  three  months'  rate,  for  example,  around  2.3  per  cent, 
which  was  about  1  per  cent  below  bank  rate.1  Deposits  in  the  nine  clearing 
banks  moved  above  1929  figures  in  the  second  half  of  the  year,  as  did 
investments  and  discounts.  Town  and  metropolitan  clearings  were  but 
moderately  smaller,  but  provincial  clearings  by  over  15  per  cent.  New 
capital  issues  for  Great  Britain  were  less  by  20  per  cent.  Prices  of  fixed- 
interest  securities  rose,  and  prices  of  industrial  stocks  fell  in  the  way  that 
is  normal  under  the  circumstances.  Professor  Bowley's  index  of  average 
weekly  wages,  which  had  declined  by  one  point  and  a  half  in  1927,  by 
two  points  in  1928,  and  by  one  in  1929,  fell  by  another  point  and  a  half, 
and  thus  only  continued  in  1930,  without  acceleration,  a  tendency  it 
had  displayed  before,  while  real  wage  rates,  of  course,  rose  considerably 
within  the  year. 

The  reader  should  judge  for  himself  how  far  this  pattern  accords  with 
our  model.  In  doing  so  he  should  observe  that  in  order  to  account  for 
the  very  great  difference  between  it  and  the  American  pattern  we  need 
not  fall  back  upon  anything  external  to  that  model.  Nor  could  we  do  so 
if  we  wished,  for  although  there  is  no  doubt,  of  course,  that  the  English 
organism  was  much  more  exposed  to  external  influences  (in  our  sense) 
than  the  American,  nothing  but  depressing  effects  could  have  emanated 
from  them.  The  most  important  has  already  been  mentioned.  Money 
value  of  total  exports  fell  off  by  approximately  one-fifth  (1929,  720 
million  pounds;  1930,  571;  peak  1924,  801),  that  of  all  manufactures  by 
about  one-fourth,  the  greater  part  of  this  loss  being  concentrated  on  the 
second  half  of  the  year.  Income  from  foreign  investments  also  declined. 
Domestic  policy  may  for  the  purpose  in  hand  be  described  as  neutral. 
The  slum  clearance  and  the  coal  acts  exerted,  the  first  no  effect,  the 
second  little  effect  during  the  year.2  The  budget  for  1930-1931— it 
was  Mr.  Snowden's  second  budget,  though  the  first  for  which  he  took 
full  responsibility — provided  for  a  considerable  increase  in  ordinary 
expenditure,  most  of  which  was,  however,  due  to  the  "derating  plan" 
of  1929-1930,  i.e.,  the  transfer  of  certain  local  burdens  to  the  national 
budget,  which  only  then  took  full  effect.  Income  tax  was  increased  by 
sixpence,  allowances  for  smaller  incomes  being  so  adjusted  that  the 
increase  was  effective  for  only  about  one  quarter  of  all  income  tax  payers, 

1  Bank  rate  was  reduced  to  3^  per  cent  on  Mar.  20,  1930.     The  maximum  of  6^>  per 
cent  occurred,  it  will  be  remembered,  Sept.  26,  1929-Oct.  31,  1929.     At  the  beginning  of 
1930  it  was  5,  from  Feb.  6,  4^;  and  from  Mar.  6,  it  was  4  per  cent.     Again,  as  in  the  case 
of  the  United  States,  it  is  assumed  that  the  question  of  the  influence  of  1929  rates  has  been 
disposed  of  in  the  preceding  chapter,  Sec.  F.     The  gold  situation,  which  did  not  grow  worse 
until  November,  will  be  more  conveniently  dealt  with  later  on. 

2  Average  net  selling  value  of  coal  at  pit  head  and  wages  per  man  shift  were,  however, 
both  slightly  higher  at  the  end  of  December  1930  than  at  the  end  of  December  1929. 


920  BUSINESS  CYCLES 

and  the  surtax  and  inheritance  tax  were  raised,  the  additional  burden 
being  considerable  only  in  the  higher  brackets.  Without  receding  from 
his  general  opinion  about  the  effects  of  fiscal  policies  of  this  type,  the 
writer  believes  that  the  increments  decided  on  cannot  have  accentuated 
the  depression  materially  and  that,  on  the  contrary,  the  budget  may  have 
strengthened  confidence  in  the  seriousness  of  labor  finance.1  All  this 
was  not  "pump  priming."  But  neither  was  it  "deflation." 

3.  Germany. — Though  German  developments  during  that  year, 
unlike  American,  but  like  English  developments,  could  n&t  be  com- 
pletely understood  without  appeal  to  external  factors,  fundamental 
contours  nevertheless  answer  to  our  schema.  To  factors  external 
to  the  economic  process  we  attribute,  as  has  been  explained  in  Chap. 
XIV,  Sec.  E,  the  early  emergence  of  difficulties  right  after  the  strong 
prosperity  of  1927.  German  industrial  production  and  employment 
fell  already  in  1928 — more  precisely  from  the  last  quarter  of  1927 — and 
the  former  but  weakly  indicates  the  bulge  in  1929,  which  is  so  marked  in 
English  and  American  graphs,  while  unemployment  in  that  year  was 
getting  nearly  as  serious  as  it  had  been  in  1926.  Prices  of  shares  had 
declined — this,  of  course,  was  perfectly  normal,  much  more  so  than  the 
American  boom — ever  since  April  1927,  and  so  had  issues  of  fixed  interest 
securities — this  was  contrary  to  what  we  should  expect  to  happen  within 
our  process  in  a  Juglar  recession,  though  readily  explainable  under  the 
circumstances — and  government  as  well  as  municipal  finances  had  begun 
to  be  a  source  of  anxiety  in  1928.  It  is  thus  not  surprising  that  people 
were  talking  about  a  "crisis"  from  the  beginning  of  1929,  general  uneasi- 
ness being  intensified  by  failures  of  the  type  that  reveals  unsound  financial 
practice  (Frankfort  Insurance  Co.,  August  1929,  the  failure  of  which  had 
nothing  to  do  with  insurance  but  all  the  more  with  promotion — the 
German  Hatry  case),  by  the  acceptance  of  the  Young  plan,2  by  the 
sudden  withdrawal  of  foreign  balances  in  April  and  May  1929,  by  com- 
munist unrest  issuing  in  an  outbreak  in  the  latter  month,  by  the  swelling 

1  The  deficit  eventually  amounted,  as  per  accounts,  to  84  millions,  but  disappears 
when  properly  corrected  for  items  which  should  not  be  counted  in  figuring  out  a  deficit. 
On  Consolidated  Revenue  and  Expenditure  Account  (cf.  Colin  Clark,  op.  cit.,  p.  140) 
of  Government,  Local  Authorities,  and  Social  Insurance,  there  was  an  excess  of  Income 
(1,022  million  pounds)  over  Expenditure  (982  million  pounds).  Unless  we  assume  that 
all  of  that  billion  was  paid  out  of  existing  and  active  deposits,  we  are  driven  to  the  conclu- 
sion that  public  expenditure  added,  though  only  a  little,  to  the  sum  total  of  income,  i.e.* 
that  there  was  a  small  amount  of  "income  generation"  through  public  expenditure. 

8  Whatever  intentions  and  the  merits  of  the  case  may  have  been,  the  Young  plan  sapped 
the  political  strength  of  the  government  and  the  parties  responsible  for  its  acceptance 
much  more  than  they  themselves  at  first  realized.  Though  it  did  not  immediately  create 
any  difficulties,  and  though  it  fully  provided  for  all  difficulties  which  were  thought  likely 
to  arise,  it  contributed — though  under  rather  than  on  the  surface — to  that  state  of  the 
public  mind  and  is  thus  relevant  to  our  subject. 


THE  WORLD  CRISIS  AND  AFTER  921 

f  social  burdens,  by  the  inability  or  unwillingness  of  government  to 
nth  the  labor  situation,  and  by  a  spreading  sense  of  the  brittleness 

political  structure  of  the  time. 

hile  objective  facts  do  not  quite  bear  out  that  pessimistic  attitude 
129,  the  situation  deteriorated  in  1930  more  rapidly  in  Germany 
in  England  or  the  United  States.  It  is  true  that  public  spending 
»elow)  kept  up  consumption  and  that  sales  of  cooperatives — even 
il  retail  trade  in  such  articles  as  furniture — fell  off  but  little.  It  is 
rue  that  hourly  wage  rates  did  not  fall  till  the  last  months  of  1930, 
i  the  contrary  slightly  rose  until  then,  keeping  on  a  level  of  nearly 
•  cent  above  1925  and,  for  the  year,  also  above  1929.  But  this  does 
tiaracterize  the  situation,  because  they  were  so  kept  by  political, 
and  administrative  pressure,  particularly  by  the  practice  developed 
e  official  arbitrators  (Schlichter),  and  because,  these  official  rates 

trade  union  minima,  the  rates  actually  paid  may  have  fallen 
ut  affecting  wage  statistics,  provided  they  still  remained  above 
minima.1  According  to  an  investigation  of  the  Verband  Deutscher 
lindustrieller  for  the  Berlin  district,  the  latter,  however,  does  not 
to  have  been  the  case,  at  least  to  an  appreciable  extent,  until  the 
in  of  1930.  No  doubt,  of  course,  can  be  entertained  about  the 
,se  of  real  wages,  the  Reich's  cost  of  living  index  (1913-1914  =  100) 
$  at  an  accelerating  rate  from  151.6  in  January  to  141.6  in  December, 
number  of  symptoms  behaved  in  a  genuinely  normal  way.  Bond 

rose  until  July — the  London  quotation  of  the  Dawes  loan  began  to 

May — when  the  political  situation  provided  a  reason  for  a  down- 
money  rates  fell  till  about  August,  the  bank  rate  staying  at  4  per 
the  postinflation  minimum,  from  June  to  August;  Reichsbank 
igs  and  debits  to  postal  accounts  receded  moderately;  advances  in 
it  account  (Schuldner  in  laufender  Rechnung;  Grossbanken  only) 
,sed  and  so  did  the  amount  of  bills  held  by  banks,  though  under 
tions  of  monetary  ease  the  Reichsbank's  portfolio  decreased  sharply 
I  the  first  three  quarters  of  the  year. 

le  gold  withdrawals  from  middle  of  September  to  middle  of  October, 
nting  to  about  1  billion  marks,  were  met  and  the  reserve  was 
dshed  to  the  extent  of  half  of  that  sum  from  the  proceeds  of  a  foreign 
^ency  loan  negotiated  by  the  federal  government.  They  were 
Y  political  in  nature.2  The  increase  in  bank  rate  which  they  induced 

so,  under  conditions  of  severe  unemployment,  actual  payments  below  those  minima 
>ssibly  have  been  connived  at  by  trade  unions.  The  "rigidity  of  wages"  thus  may 
ious  in  part. 

is  just  possible  that  small  withdrawals  would  have  been  made  in  any  case,  owing  to 
ceding  fall  in  German  money  rates.  But  since  in  other  countries  rates  had  fallen  as 
>r  more,  this  is  not  likely.  Nor  was  it  likely,  circumstances  being  what  they  were, 


922  BUSINESS  CYCLES 

led  to  a  stiffening  of  rates  all  round,  but  the  situation  being  what  it  was, 
an  increase  of  lj^  per  cent  (prime  bankers'  acceptances)  can  hardly  have 
been  a  major  consideration  in  short-term  operations,  while  nobody  was 
ready  to  embark  upon  long-time  investment  in  any  case. 

Total  industrial  production  shrank,  relatively,  as  much  or  more  than 
it  did  in  this  country,  and  it  did  so  from  the  very  beginning  of  the  year. 
The  January  figures  for  crude  steel  and  products  of  rolling  mills  were 
down  by  about  13  and  10  per  cent  as  compared  with  the  "figures  for 
January  1929,  which  were  already  low.1  By  December  the  (imperfect) 
index  of  production  landed  at  72.5  per  cent  of  1928.  The  total  number  of 
unemployed  was  about  3  millions  during  the  first  10  months  and  rose  to 
nearly  4.4  in  December.  There  cannot  be  any  doubt  that  the  extra- 
economic  factors  alluded  to  had  much  to  do  with  this.  Superimposing 
themselves  on  what  would  in  any  case  have  been  a  depression,  they 
intensified  it  greatly,  as  some  of  them  had  previously  damped  prosperi- 
ties. The  writer  does  not  know  of  any  facts  not  covered  by  this  diagnosis. 
In  particular,  it  is  (when  we  recall  what  those  factors  were)  reasonable  to 
say  that  they  could  not  have  produced  a  slump  of  that  magnitude  by 
themselves,  i.e.,  if  the  organism  had  been  in  one  of  its  positive  phases. 

Government  was  influenced  by  two  opposite  considerations  or,  rather, 
was  under  pressure  from  two  different  camps.  Many  interests,  especially 
the  agrarian  interest,  the  condition  of  which  grew  steadily  worse,  and  the 
labor  interest,  which  anticipated  a  breakdown  of  the  unemployment 
insurance  and  further  increase  in  unemployment,  needed  and  pressed  for 
help.  There  was  also  increasing  demand  for  pump  priming.  Other 
interests  pointed  and  pressed  in  the  direction  of  an  attempt  to  use  the 
depression  in  order  to  normalize  general  conditions,  i.e.,  to  break  with  the 
spending  habits  of  public  bodies,  to  retrace  steps  in  social  legislation,  to 
reduce  money  wage  rates,  to  buttress  the  currency  and  its  purchasing 
power,  and  so  on.  Such  views  were  not,  however,  held  by  interested 
parties  only.  Many  people  even  believed  not  only  that  such  a  policy 
would  provide  sound  foundations  for  future  progress,  but  also  that  it 
would  have  remedial  effects  at  the  moment.  "In  the  years  from  1927 
to  1929  we  have  been  wandering  in  an  economic,  financial,  and  socio- 
political labyrinth  (or  maze;  the  German  word  was  Irrgarten) .  And 
now  we  must  return  to  sober  reality" — so  the  federal  minister  of  labor, 

that  an  increase  in  bank  rate  of  one  per  cent  would  exert  any  protecting  or  attracting 
influence.  As  far  as  that  goes,  the  increase  was  merely  ceremonial.  And  since  there  was 
no  excessive  activity  to  restrain,  it  is  difficult  to  see  what  purpose  it  was  to  serve.  Econo- 
mists imbued  with  a  mystical  belief  in  the  efficacy  of  interest  rates  will  no  doubt  explain 
all  that  was  to  follow  by  it.  This  cannot  be  helped. 

1  Coal  output  was,  however,  higher  for  January  and  February,  output  of  coke  for 
January. 


THE  WORLD  CRISIS  AND  AFTER  923 

Stegerwald,  himself  a  prominent  labor  leader,1  declared  as  late  as  Decem- 
ber 1931.  Whatever  the  merits  or  demerits  of  such  a  policy,  which 
has  been  and  is  so  unintelligently  recommended  by  some  and  so  unin- 
telligently  rejected  by  others,  it  is  important  for  our  purpose  to  ask  how 
far  it  was  proceeded  with  during  that  critical  year.  In  doing  so  we  must 
realize  that  those  two  policies  were  not  so  mutually  exclusive  as  they  might 
seem.  Certain  steps  on  the  lines  of  the  first  were  hardly  avoidable 
precisely  in  order  to  get  maneuvering  space  for  the  second.  Some  steps 
on  the  lines  of  the  second  even  an  adherent  of  the  first  would  have  had  to 
take  if  he  was  to  avoid  chaos  in  a  country  which  had  had  so  recent  an 
experience  of  what  inflation  is.  A  horseman  who  alternately  collects 
and  gives  to  his  mount  does  not  thereby  contradict  himself  or  prove  his 
lack  of  purpose. 

The  government  spent  freely.  The  federal  deficit  of  over  1  billion 
marks  (1930-1931)  following  upon  a  deficit  of  712  millions  (1929-1930) 
was  relatively  more  income  generating  than  was  the  deficit  of  the  United 
States.  In  taking  measures  in  aid  of  agriculture,  in  providing  means  for 
those  unemployed  whom  the  statutory  or  financial  limits  of  unemploy- 
ment insurance  failed  to  include,  in  doing  some  pump  priming  and 
straining  its  credit  (the  slogans  were  cranking,  Ankurbelung,  and  bridging 
credits,  Ueberbruckungskredite),  the  government  certainly  discovered  and 
acted  upon  the  theory  of  the  unbalanced  budget.  By  the  (amendment 
of  the)  bank  act  of  Mar.  13,  1930,  long-term  bonds  of  the  Reich,  of  the 
states,  and  of  the  municipalities  became  eligible  as  collateral  for  Reichs- 
bank  credits,  and  toward  the  end  of  the  year  treasury  bills  were  freely 
discounted  by  the  Reichsbank.  This  is  not  exactly  " deflation,"  whatever 
that  term  may  mean.  Two  moves  on  the  other  line  were  obviously 
intended  to  serve  as  safeguards  in  the  long  run.  First,  the  shadow  of  a 
curb  was  provided  for  future  excesses  of  municipal  finance  by  certain 
regulations  intended  to  make  it  more  difficult  for  municipalities  to  run 
into  debt  (Dec.  3,  1930).  Second,  the  Reichsbank  nailed  its  colors  to  the 
mast  by  putting  into  operation  Clause  31  of  the  bank  act  of  1924,  which 
bound  it  to  do  what,  so  far,  it  had  done  without  obligation,  viz.,  to 
redeem  its  notes  in  gold  or  exchange  (Apr.  19).  Both  measures  might 
have  become  effective  in  the  future,  but  certainly  cannot  have  had  effects 
in  1930. 

The  only  other  point  that  matters  here  was  the  government's  move  to 
reduce  rigid  prices  (end  of  August),  which  went  parallel  to  its  efforts  to 
keep  up  agrarian  prices.  The  index  of  prices  at  wholesale  (Institut  fur 

1  He  was  not  a  socialist.  But  socialist  labor  leaders  did  not  object  as  strongly  as  they 
assuredly  would  have  done  if  they  had  really  thought  him  wrong.  The  writer's  impression 
is  that  they  too  were  of  the  opinion  that  things  had  gone  too  far  and  rather  felt  relieved 
when  somebody  else  shouldered  the  unpopularity  of  leading  back  to  the  highway. 


924  BUSINESS  CYCLES 

Konjunkturforschung:  1913  =  100)  for  1930  was  114,  as  against  131  for 
1929,  and  fell  rapidly  within  the  year.  But  the  government  and  the 
public  were  as  much  exercised  about  "administered"  prices  as  the 
American  government  and  public  are  now.  Moreover,  the  argument 
presumably  suggested  itself  that  it  would  be  easier  to  reduce  wage  rates 
if  the  government  aimed  also  at  other  points  in  the  price  structure. 
Accordingly,  it  embarked  upon  a  campaign  for  the  reduction  of  prices  and 
costs,  while  the  Reichsbank  busied  itself  until  October  in  bringing  about  a 
general  reduction  in  interest  rates,  particularly  in  East  Prussia.1  We  are 
not  concerned  with  the  logic  of  all  this.  Since  it  worked  with  and  not 
against  the  stream,  it  probably  effected  little  beyond  somewhat  quicken- 
ing adaptation  in  sticky  spots.  A  number  of  rigid  raw-material  prices, 
which  had  not  moved  at  all  during  1929  and  the  first  months  of  1930, 
fell  by  a  little  over  10  per  cent  toward  the  end  of  the  year.  The  weighted 
average  of  trade  union  wage  rates  declined  by  5.8  per  cent  in  the  first 
4  months  of  1931,  while  the  index  of  production  increased  by  9.1  per  cent 
in  the  first  quarter,  and  employment  by  5  per  cent  during  April  and  May 
1931,  both  indices  being  cleared  of  seasonals.  Presence  of  a  causal  rela- 
tion was  claimed,  possibly  not  quite  without  foundation.2 

C.  1931  and  1932. — While  it  was  necessary  to  have  the  facts  of  1930 
firmly  planted  in  the  reader's  mind,  it  seems  possible  to  confine  discussion 
of  the  two  years  which  span  the  real  "catastrophe" — of  the  capitalist 
system  or  of  an  individual  system  of  economic  values — to  a  number  of 
comments  that  can  easily  be  worked  up  into  a  connected  survey.3 

1.  Physical  Production. — The  fundamental  point  to  emphasize  is 
again  that  there  is  nothing  in  the  fact  that  severely  depressive  symptoms 

1  Annual  report  to  the  shareholders'  meeting,  Apr.  29,  1931. 

2  So,  for  example,  in  Lohnpolitische  Kurvenbilder  zur  Krisenlage,  Verein  Deutscher 
Maschinenbau-Anstalten  1931   (mim.).     Even  the  "possibility"  that  that  argument  be 
"not  quite"  without  foundation  really  goes  much  further  than  most  modern  economists 
would  care  to  go.     The  present  writer  would  not  go  further  than  that.     But  that  possibility 
he  is  prepared  to  prove.     That  pamphlet  does  not  fail  to  notice  the  lag  between  increase  of 
output  and  increase  of  employment,  which  it  explains  partly  on  statistical  grounds  and 
partly  on  the  ground  that  in  such  situations  the  labor  employed  is  underutilized. 

8  There  is  some  danger  in  thus  leaving  room  for  uncertainty  of  interpretation  on  many 
points  which  could  be  removed  only  by  thorough  discussion  of  every  detail.  But  we  have 
no  choice.  Some  general  impressions  can  be  gleaned  from  referring  to  our  charts.  The 
current  reports  in  the  Review  of  Economic  Statistics  (W.  L.  Crum)  and  the  London  and 
Cambridge  Service  would  prove  helpful.  Beyond  this,  the  simplest  way  of  testing  and 
supplementing  the  statements  of  our  text  is  perusal  of  the  World  Economic  Surveys  by 
Professor  J.  B.  Condliffe  (League  of  Nations,  Economic  Intelligence  Service).  Professor 
Bertil  Ohlin's  Course  and  Phases  of  the  World  Economic  Depression  will  prove  a  valuable 
guide  up  to  the  summer  of  1931.  Comparison  between  those  interpretations  and  the  one 
offered  here  is  invited.  Mention  should  be  made,  also,  of  Professor  Lionel  Robbins's 
analysis  in  The  Great  Depression. 


THE  WORLD  CRISIS  AND  AFTER  925 

continued  to  dominate  the  picture  in  all  three  countries  to  require  addi- 
tional explanation  from  our  standpoint — nothing  has  to  be  explained 
away,  and  that  fact  fully  conforms  to  expectation  from  our  model.  As 
stated  in  Sec.  A,  that  picture  reflects  ever  since,  roughly  speaking,  the 
middle  of  1930  an  element  not  in  evidence  before — the  "vicious  spiral" 
which  for  nearly  two  years  would  by  itself  suffice  to  describe  the  surface  of 
what  was  going  on.  But  we  know  that  the  process  thus  designated  is  to 
a  greater  or  lesser  extent  a  feature  of  any  depression.  It  is,  in  fact,  largely 
responsible  for  turning  the  "normal  liquidation"  of  recession  into  the 
"abnormal  liquidation"  of  depression  (Chap.  IV).  At  that  stage>  the 
phenomena  expressed  by  the  "principle  of  cumulation  or  acceleration" 
and  by  the  "debt-deflation"  theory  become  part  of  the  cyclical  mecha- 
nism, a  particularly  important  part  when  all  three  cycles  are  in  the  phases 
most  favorable  to  them.1  But  since  we  also  know  that  the  working  of 
the  spiral  is  erratic  ("internally  irregular")  and  extremely  sensitive  to 
incidents,  accidents,  and  external  factors,  and  that  the  troughs  it  creates 
are  intrinsically  unreliable,  we  must,  especially  in  the  face  of  the  various 
recovery  policies  that  were  taking  shape  in  those  two  years,  recognize  the 
limitations  of  any  attempt  at  verification  beyond  that  fundamental 
fact.  Barring  crises  and  panics,  which  may  occur  at  any  time  and  the 
occurrence  and  effects  of  which  have  simply  to  be  registered,  we  expect  a 
depression  to  display  shrinkage  at  decreasing  rates  until  it  shades  off  into 
recovery,  in  the  case  of  the  Kondratieff  on  a  very  broad  bottom  and  in  the 
case  of  the  Juglar  on  a  bottom  of  about  a  year,  while  the  Kitchin  phases 
should  show  in  what  statisticians  call  surface  movements  in  series  cor- 
rected for  seasonals.  Even  without  attaching  weight  to  the  dates  yielded 
by  our  experimental  count  (see  above,  Sec.  A)  we  will  note  that  since  a 
Juglar  depression  must  contain  positive  Kitchin  phases — and,  if  we  trust 
our  schema  to  that  extent,  end  up  with  a  Kitchin  prosperity — it  should 
display  an  observable,  though  possibly  weak  and  short,  upturn  at  the 
bottom,  which  again  is  nothing  else  but  our  way  of  expressing  a  familiar 
fact  of  business  experience. 

1  Hence,  it  would  not  constitute  valid  objection  to  say  that  the  events  of  these  years  do 
not  require  explanation  by  the  theory  of  innovation  but  are  adequately  explained  by  the 
vicious  spiral  or  by  the  principle  of  acceleration.  The  one  and  the  others  do  not  stand  on 
the  same  plane  of  argument  and  cannot  be  pitted  against  each  other.  Believing  he  has 
made  this  abundantly  clear  throughout  this  book,  the  writer  wishes  to  add  only  two  minor 
points.  First,  the  word  acceleration  seems  to  suggest  increasing  rates  of  change;  this 
would,  however,  even  if  the  word  were  to  be  taken  at  face  value,  not  contradict  the  expecta- 
tion to  be  presently  met  with  in  the  text,  because  that  word  refers  only  to  a  component 
and  the  expectation  to  a  resultant.  Second,  it  might  be  held  that,  the  facts  covered  by  the 
term  Vicious  Spiral  being  the  outstanding  feature  of  those  years,  we  may  stop  at  that  and 
discuss  the  theoretical  and  practical  problems  of  the  situation  without  going  beyond  those 
facts.  This  is  not  so.  It  is  not  indifferent,  for  either  diagnosis  or  therapy,  what  starts  the 
spiral. 


926  BUSINESS  CYCLES 

What  are  the  facts?  According  to  a  very  general  opinion  with  which 
we  agree  in  this  case,  a  preliminary  answer  may  be  derived  from  the 
behavior  of  industrial  production:  nothing  else  qualifies  equally  well  for 
the  role  of  an  indicator  of  the  objective  state  of  the  system  in  the  later 
part  of  depression  and  during  recovery.  We  will  look  first  for  the  lower 
turning  point,  or  rather,  remembering  our  view  about  troughs,  for  the 
lowest  segment  of  our — or  almost  any1 — graphs.  In  almost  all  countries, 
however  different  their  structures  and  general  conditions,2  particularly  in 
Austria,  Belgium,  France,  Germany,  Hungary,  Italy,  Poland,  and 
Sweden,  it  occurs  in  (the  middle  of)  1932.  In  the  Canadian  index  the 
trough  comes  in  February  1933.  Japan  was  an  exception  all  along,  in 
fact,  the  standard  instance  for  advocates  of  inflationary  policies,  who 
have  only  to  be  reminded  that  the  case  is  a  special  one  in  that  this  policy 
found  all  the  conditions  ready  at  hand  for  a  rapid  industrialization  of  the 
country.  In  the  English  case  there  is  some  doubt,  which  will,  however, 
be  presently  removed;3  but  our  chart  displays  a  well-marked  trough  in  the 
summer  of  1932.  So  it  does  for  this  country.  The  relapse  in  the  spring 
of  1933,  which  carried  the  American  index  back  to  about  its  previous  low, 
while  not  in  itself  astonishing — that  after  a  depression  of  such  severity 
relapses  should  occur  in  the  first  stages  of  recovery  is  perfectly  under- 
standable on  common-sense  grounds,  even  without  appeal  to  the  course 
of  Kitchin  phases — was  serious  enough  to  suggest  the  presence  of  dis- 
turbing factors.  But  since  the  advent  of  a  new  administration  pledged 
to  pursue  an  active  policy  and  the  third  epidemic  among  banks — which 
will  be  dealt  with  below — readily  supply  the  explanation,  it  seems  reason- 

1  Graphs  displaying  the  behavior  of  indices  of  production  in  different  countries  are,  in 
spite  of  the  problems  that  arise  on  the  score  of  comparability,  among  the  most  common 
tools  of  scientific  and  popular  analysis  of  the  crisis.     Specific  mention  should,  however, 
be  made  of  the  work  of  the  National  Bureau — Professor  Mills  in  the  Bulletin  for  Feb.  20, 
1933,  and  Mr.  Bliss  in  the  Bulletins  for  June  26,  1934,  and  Nov.  15,  1935— and  of  Miss  D. 
Westcott,  Review  of  Economic  Statistics,  Dec.  15,  1934.     Also  see  again  N.  J.  Wall,  Monthly 
Index  of  World  Industrial  Production,  1920-1935,  Bureau  of  Agricultural  Economics,  1936. 

2  Unfortunately,  it  is  not  only  the  economic  structures  and  conditions  but  also  the 
indices  which  are  different  to  the  point  of  incomparability.     It  is  believed,  however,  that  the 
validity  of  our  argument  is  not  substantially  impaired  thereby.     The  shapes  of  the  bottom 
parts  of  the  curves  are  not  alike  and  raise  various  questions  of  detail,  which  will,  however, 
be  dealt  with  only  for  the  United  Kingdom  and  the  United  States.     If  we  can  trust  the 
index  or  the  available  indications,  Rumania  and  Spain  did  not  conform.     But  the  course 
of  things  in  Spain  was  clearly  politically  conditioned  (proclamation  of  republic,  April 
1931).    In  Tchechoslovakia  the  bottom  comprises  1932  and  1933,  but  the  steepest  descent 
was  in  the  last  quarter  of  1931  and  the  first  quarter  of  1932,  after  which  the  index  crawled 
along  level  until  late  1933. 

8  In  the  London  and  Cambridge  Service  annual  index  (including  agriculture  and  build- 
ing) the  low  occurs  for  1931,  but  1932  displays  the  annual  minima  in  iron  and  steel,  non- 
ferrous  metals,  food,  drink,  tobacco,  and  the  leather  group.  In  mining  1933,  in  textiles 
1930  is  the  minimum  year. 


THE  WORLD  CRISIS  AND  AFTER  927 

able  to  accept  the  1932  trough  as  the  "true"  one,  although  according  to 
some  indices — not,  however,  that  of  the  Federal  Reserve  Board — the 
trough  in  March  1933  is  somewhat  deeper  still.  For  both  manufactures 
and  minerals,  also  for  steel,  lumber,  petroleum  refining,  coke  (by-product), 
foods  (the  descent  of  which,  however,  was  hardly  perceptible)  textiles, 
automobiles,  and  construction  (values)  taken  separately,  decline  came  to 
a  stop  by  roughly  the  middle  of  1932.  Only  for  cement,  rubber  products, 
and  power  production,1  possibly  also  tobacco,  the  halt  does  not  come 
before  the  spring  of  1933,  while  leather  and  leather  products  turned  up  by 
the  end  of  1931.  The  behavior  of  carloadings  and  imports  (current 
values)  bear  out  the  general  picture  and  so  did  business  failures,  which 
fell  sharply  after  reaching  their  peak  in  the  summer  of  1932.  The 
Federal  Reserve  Board's  index  of  department-store  sales  declined  until 
the  spring  of  1933,  but  this  was  due  to  the  continued  fall  in  prices. 

Thus  it  seems  that,  so  far  as  physical  production  is  concerned,  the 
location  of  the  bottom  of  the  depression  answers  well  to  our  idea  as  to 
where  it  should  have  occurred.  Turning,  in  the  second  place,  to  the 
descent  to  that  bottom,  we  find2  that  production  in  the  United  Kingdom 
suffered  its  most  severe  contraction  in  1930,  and  that  the  rate  of  shrinkage 
was — even  disregarding  the  little  hump  between  the  autumn  of  1931  and 
the  autumn  of  1932 — smaller  for  1931  and  still  smaller  in  1932,  the  move- 
ment retaining  its  comparative  mildness  throughout.  A  smooth  curve 
fitted  to  the  German  figures  from  the  middle  of  1930  to  the  end  of  1932 
would  display  a  monotonically  decreasing  rate  of  fall,  shading  off  into 
increase.  American  industrial  production  declined  for  1932  as  a  whole 
very  much  less  than  it  did  in  1931  and,  owing  to  the  upturn  at  the  end, 
the  average  rate  of  change  within  the  year  would  also  be  smaller.  But 
the  contraction  at  the  beginning  of  1932  was  (in  percentage  rate)  the  most 
serious  of  all,  and  whether  the  decline  of  1931  was  milder  than  that  in  the 
second  half  of  1930  is  a  matter  of  index  construction.3  These  irregular- 

1  Electrical  power  production  did,  however,  strike  a  plateau  during  the  two  middle 
quarters  of  1982,  before  it  took  another  dive.     As  has  been  mentioned,  sales  for  domestic 
use  continued  to  increase  even  in  1932. 

2  See,  for  example,  D.  Westcott,  op.  cit.,  p.  256,  or  our  own  Chart  XLII.     For  Germany 
Miss  Westcott's  curve  is  better  than  ours.     The  total  decline  from  the  1929  peak  to  the 
trough  was,  of  course,  much  greater  in  the  United  States  than  in  most  countries.     This  is 
certain,  in  spite  of  all  the  doubts  about  comparability.     Mr.  Bliss  (07?.  cit.,  p.  2)  estimates 
it  at  53  per  cent,  while  for  Germany  the  corresponding  figure  is  43  and  for  the  United 
Kingdom  (on  quarterly  data)  22  per  cent.     We  will  add  the  percentage  fall  of  crude  steel 
production  for  the  first  9  months  of  1932  as  compared  with  1929:  Germany,  66.4;  Great 
Britain,  46.3;  United  States,  76.5. 

3  According  to  the  index  chosen  for  our  chart,  decline  for  the  first  half  of  1930  was  no 
less  precipitous.     But  this  is  clearly  out  of  keeping  with  the  general  aspect  of  business  at 
that  time. 


928  BUSINESS  CYCLES 

ities  may,  however,  be  partly  accounted  for  by  the  specifically  American 
difficulties  in  the  sphere  of  credit  and  banking. 

We  cannot  stay  to  note  various  other  points  of  interest.1  One  fact 
only  which  sheds  much  light  on  the  nature  of  the  process  cannot  be  passed 
by,  viz.y  that  the  depression  acted  as  an  efficiency  expert.  This  holds 
true  of  all  three  (and  other)  countries,  but  it  is  particularly  in  evidence  in 
the  United  States.  Output  per  man-hour,  which  may  be  roughly  said  to 
have  increased  by  22  per  cent  from  1923  to  1929,  surpassed  in  1932  the 
1929  figure  by  an  amount  that  will  greatly  vary  according  to  the  indices 
entering  into  the  computation,  but  may  plausibly  be  put  in  the  neighbor- 
hood of  20  per  cent.2  This  was,  of  course,  not  only  due  to  rationalization 
under  duress  in  the  concerns  that  continued  operations,  but  also  to  the 
shutting  down  or  permanent  elimination  of  others,  which  may  as  a 
broad  rule — though,  as  we  know,  by  no  means  always — be  assumed  to 
have  been  less  efficient.  The  second  component,  however,  enters  our 
model  no  less  than  the  first.  More  rigorous  selection  of  the  work- 
men employed  or  to  be  employed  may  also  have  had  some,  the  mere 
underutilization  of  fixed  factors  cannot  have  had  significant  effect — the 
latter  factor  is  likely,  under  the  conditions  of  modern  industry,  to  have 
worked  the  other  way,  at  least  in  many  cases.  Reference  to  Chart 
XLIII  is  all  that  is  necessary  as  regards  the  behavior  of  group  indices. 
It  was  as  we  should  expect,  foods  and  perishable  commodities  in  general 
keeping  up  comparatively  well  throughout  the  bottom  of  the  depression.3 

For  most  countries  total  industrial  as  well  as  manufacturing  produc- 
tion experienced  a  beginning  of  recovery,  as  small  as  we  should  expect, 

1  See,  for  example,  Charles  A.  Bliss,  op.  cit.,  p.  10,  for  the  wide  dispersion  of  percentage 
changes  1929  to  1932  in  individual  items.     Note  the  increase  in  refrigerators  and  rayon  and 
the  particularly  sharp  decrease  in  railroad  cars,  locomotives,  pig  iron  and  iron  ore. 

2  Professor  Mills  (op.  cit.,  p.  4)  arrives  at  12  per  cent,  Mr.  Bliss  (op.  cit.t  p.  6)  at  21  per 
cent  (manufacturing  only),  which  in  the  second  publication  quoted  before  he  reduces  to 
18  per  cent.     See  the  comments  of  both.     It  is  in  accordance  with  our  view  of  the  processes 
of  negative  phases  that,  as  both  authors  state,  this  increase  was  not  due  to  any  significant 
extent  to  "revolutionary"  innovations.     But  if  both  point  to  the  decline  in  the  production 
of  machinery  between  1931  and  1932  in  support  of  the  contention  that  installation  of  new 
equipment  was  not  a  dominant  factor  in  that  increase  in  output  per  man-hour,  we  must  not 
forget  that  installation  of  new  equipment  exerts  its  influence  with  a  lag.     This  is  so  even 
within  the  individual  concern,  but  still  more  in  industry  taken  as  a  whole,  because  the 
process  of  adaptation  and  elimination  of  less  mechanized  or  otherwise  less  efficient  concerns, 
which  is  the  backbone  of  depression,  is  a  slow  process. 

8  For  details  on  the  behavior  of  household  consumption,  see  again  A.  R.  Tebbutt,  op.  cit. 
In  particular,  it  is  interesting  to  note  that  deflated  sales  of  furniture  at  department  stores 
in  1931  were  below  1929  by  no  more  than  about  12  per  cent.  Toilet  articles  and  drug 
sundries  even  kept  their  dollar  volume  substantially  intact.  Cigarette  consumption  did 
not  decline  substantially  before  1932,  when  it  still  kept  about  the  1928  level.  Gasoline 
consumption  rose  in  1931  and  fell  in  1932  by  only  10  per  cent. 


THE  WORLD  CRISIS  AND  AFTER  929 

in  the  second  half  of  1932,1  which  slackened  toward  the  end  of  the  year, 
when,  as  we  would  say,  the  last  Kitchin  of  the  Juglar  turned  into  reces- 
sion. Such  a  relapse  at  that  juncture  is  in  itself  no  problem.  The 
majority  of  countries  for  which  indices  are  available  display  it,  but  not 
Germany — nor  France,  where,  however,  industrial  production  embarks 
upon  a  decline  in  the  middle  of  1933 — whose  production  moves  up  steadily 
and  strongly  from  the  third  quarter  of  1932  to  the  middle  of  1934.  In 
the  United  States  production  of  manufactures  and  mining  as  measured 
by  the  seasonally  adjusted  index  of  the  Federal  Reserve  Board  increased 
substantially  during  August  and  September — from  the  July  low  of  58  per 
cent  of  the  1923  to  1925  average  to  66  per  cent — and  went  on  at  that  level 
in  October;  then,  after  a  slight  setback,  regained  it  in  December.  It  is 
important  to  note  that  it  was  the  activity  of  cotton,  woolen,  silk,  rayon 
mills,  shoe  factories,  and  the  like  which  was  responsible  for  the  increase  in 
August,  while  automobile  production  still  declined  and  steel  and  lumber 
industries  did  not  show  even  seasonal  advance.  In  September,  though 
the  steel  industry  expanded  a  little — it  then  reached  20  per  cent  of  capac- 
ity during  the  first  three  weeks  of  October — the  same  general  features 
characterize  the  situation.  In  October,  automobile  production  increased, 
but  there  was  still  little  activity  in  investment  industries.  This  is  in 
accordance  with  the  usual  pattern.  We  know  that  in  a  four-phase  cycle 
recovery  from  the  lower  turning  point  does  not  typically  start  either  by 
innovations  or,  more  generally,  investment,  or  by  firms'  borrowing  but 
exclusively  by  moves,  from  an  indefinite  number  of  points  and  within  the 
existing  framework  of  plant,  equipment,  and  balances,  toward  a  neighbor- 
hood of  equilibrium. 

It  remains  to  account  for  the  upturn  which  occurred  in  the  industrial 
production  of  this  country  and  of  Germany,  but  not  in  that  of  the  United 
Kingdom,  at  the  beginning  of  1931.  In  this  country  more  than  in  others 
stock  prices,  pay  rolls,  loans,  velocity  of  deposits,  and  department-store 
sales  increased  along  with  production,  while  employment  at  least  ceased 
to  fall  for  over  4  months.  Readers  may  well  feel  that  we  need  not  bother 
about  this  intermezzo,  for  it  is  easy  to  understand  that  so  long  a  decline 
should  be  interrupted  by  temporary  rallies.  Optimistic  anticipations  of 
people  about  "recovery  being  around  the  corner"  may  also  count  for 
something  in  the  explanation  of  a  phenomenon  of  this  order  of  magnitude. 
But  there  may  be  more  to  it.  No  recovery  was  due  as  yet,  but  a  pro- 
gressive abatement  of  depressive  symptoms  was.  The  elimination,  by 

xThe  indices  of  world  production  of  the  League  of  Nations  (excluding  Russia)  and 
of  the  Bureau  of  Agricultural  Economics  (N.  J.  Wall,  op.  cit.,  p.  2)  turn  up  exactly  in  the 
middle  of  the  year.  So  do  the  Austrian,  Belgian,  French,  and  Swedish  indices.  In  Italy 
and  Hungary  the  upturns  occur  a  little  earlier.  The  Polish  index  rises  from  the  beginning 
of  the  year,  but  falls  to  a  new  low  in  December. 


930  BUSINESS  CYCLES 

the  processes  of  depression,  of  untenable  situations  and  of  obstacles  that 
give  way  only  under  duress,  was  partly  accomplished.  Some  adjustments 
in  the  cost  structures  of  the  kind  noticed  in  our  description  of  German 
events  of  1930  may  have  given  an  impulse  to  individual  firms  and  indus- 
tries. If  so,  subsequent  vicissitudes  would  have  to  be  explained  primarily 
in  terms  of  accidents  and  external  factors  and  the  first  half  of  1931  would, 
in  this  case,  give  us  a  fair  idea  of  what  the  rest  of  the  depression  would 
have  been  without  them. 

2.  Incidents,  Accidents,  and  Policy  in  Germany. — During  the  first 
half  of  1931  one  of  those  political  events  occurred  which,  at  least  in  their 
precise  form  and  timing,  we  cannot  but  consider  as  purely  fortuitous, 
whatever  our  theory  of  history  may  be.  Suddenly  bursting  upon  an 
unprepared  world,  the  plan  of  an  Austro-German  customs  union,  natu- 
rally interpreted  as  the  forerunner  of  political  union,  was  not,  when  pro- 
posed in  March,  met  by  either  of  the  two  methods  that  would  have, 
temporarily  at  all  events,  prevented  any  further  consequences,  viz., 
resolute  opposition  or  resolute  acceptance,  but  allowed  to  develop  into  a 
source  of  irritation  and  apprehension  until,  after  protracted  blundering 
and  floundering,  it  met  its  inglorious  end  on  Sept.  3,  which  dealt  a  fatal 
blow  to  the  prestige  of  the  last  German  cabinet  that  believed  in  action 
within  existing  treaties.  Under  ordinary  circumstances  this  affair 
would,  however,  have  remained  in  the  political  sphere  and  at  first  it 
seemed  that  it  would.  General  improvement  continued,  the  Reichsbank's 
gold  stock  increased  somewhat,  it  was  still  possible  to  borrow  abroad, 
and  even  in  the  first  days  of  May,  120  million  marks  of  stock  of  the  newly 
founded  Berlin  Power  and  Light  Company  were  placed  with  a  foreign 
syndicate.  But  then,  the  opposition  to  the  customs  union  gathering 
momentum,  foreign  balances  began  to  make  for  home — no  political  lead 
had  to  be  given  in  order  to  effect  that1 — and  this  was  bound  to  produce 
difficulties  in  the  English  and  breakdown  in  the  German  market,  which  so 
largely  relied  on  them  for  current  financing.  In  the  eight  weeks  from  the 
end  of  May  to  the  middle  of  July,  Germany  lost  at  least  £J^  billion 

1  An  additional  shock  was  imparted  by  the  suspension  of  the  Austrian  Kreditanstalt, 
the  desperate  situation  of  which  became  known  to  the  public  on  May  11,  1931,  and  only  a 
few  days  earlier  to  the  directors.  But  the  importance  of  this  event  seems  to  the  writer  to 
have  been  overemphasized.  No  doubt  the  failure  of  that  famous  institution,  though 
then  only  a  shadow  of  its  former  self,  was  apt  to  make  a  profound  impression  throughout 
the  world,  particularly  because  of  its  connection  with  the  house  of  Rothschild,  accentuated 
as  this  connection  was  by  the  fact  that  the  head  of  the  Viennese  branch  of  that  family— 
the  private  banking  house  S.  M.  v.  Rothschild  &  Co.  was  not,  however,  directly  involved — 
happened  to  be  its  president.  But  the  quantitative  importance  of  the  failure,  however 
serious  for  Austria,  was  hardly  sufficient  to  exert  any  considerable  influence  on  the  course 
of  things  in  the  world  at  large  or  even  in  Germany.  Declared  losses — swallowing  up,  it  is 
true,  both  capital  and  reserves — were  about  20  million  dollars. 


THE  WOKLD  CRISIS  AND  AFTER  931 

marks — not  counting  the  credit  of  100  million  dollars  extended  to  the 
Reichsbank  by  the  Bank  for  International  Payments,  the  Bank  of 
England,  the  Bank  of  France  and  the  Federal  Reserve  Bank  of  New  York 
on  June  25 — and  the  Reichsbank's  stock  of  gold  and  foreign  exchange 
exclusive  of  the  proceeds  of  short-term  emergency  credits  fell  dur- 
ing the  year  by  about  2,200  million  marks  to  530  millions  (middle  of 
December) . 1 

Some  aspects  of  this  situation  having  been  dealt  with  in  the  preceding 
chapter,  we  need  only  notice  the  broad  outlines  of  the  consequent 
catastrophe  of  Germany's  financial  structure,  which  spread  paralysis 
throughout  her  economic  organism  besides  destroying  a  number  of 
fundamentally  unsound  concerns.2  The  Darmstadter  and  Nationalbank 
was  closed  on  July  13  after  a  run  which  followed  upon  foreign  with- 
drawals and  the  failure  of  an  important  concern  financed  by  it,  and  a 
general  run  on  banks  ensued.  Banks  were  closed  by  government  decree 
and,  even  after  reopening,  did  not  resume  unrestricted  payments  to 
depositors  until  Aug.  5,  while  the  decree  (July  18)  "against  the  flight  of 
capital"  ushered  in  the  period  of  exchange  control,  which  was  tightened 
on  Aug.  1  and  has  since  become  ever  more  stringent  to  this  day.  Another 
leading  bank,  the  Dresdner,  had  to  be  reconstructed  by  the  federal 
government  without  actually  failing.3  Savings  banks  were  in  a  still  worse 
position  as  to  liquidity,  but  resumed  payment  on  Aug.  8.  Events  on  the 
stock  exchanges,  though  they  were  closed  from  July  13  to  Sept.  3,  were 
less  catastrophic  than  might  be  expected.  There  were  shares  which  did 
not  fall  by  more  than  10  per  cent  within  the  year,  and  after  the  shock 
imparted  by  the  devaluation  of  the  pound  there  was  some  recovery  during 
the  rest  of  the  year. 

1The  "standstill  agreement"  of  Aug.  19,  effective  from  Sept.  1,  following  upon  a 
period  during  which  Germany  discontinued  her  foreign  payments — first  without  and  then 
with  the  informal  consent  of  her  chief  creditors — of  course,  regularized  but  did  not  stop 
the  outflow  of  gold  and  foreign  exchange,  which  amounted  to  over  500  millions  in  the  last 
8  months  of  the  year.  Payment  of  foreigners'  mark  balances  and  various  lacunae  in  the 
standstill  agreement  caused  further  loss,  which  it  is  difficult  to  estimate  but  which  was 
partly  made  up  for  by  the  favorable  balance  of  trade. 

2  Losses  on  ill-conceived  ventures  did  not  greatly  affect  creditors  in  the  cases  of  the 
Karstadt  and  the  Linoleum  concerns,  but  the  case  of  the  Blumenstein  concern,  which  led 
to  default  on  the  1  million-pound  loan  of  the  Bank  ftlr  Textilindustrie,  was  more  serious. 
Still  more  so  was  the  bankruptcy  of  the  Norddeutsche  WollkUmmerei,  which  owed  240 
million  marks,  the  Darmstaedter  Bank  being  the  chief  sufferer.     This  bankruptcy  occurred 
in  June,  as  did  the  difficulties  of  the  Nordstern  insurance  company,  which  was,  however, 
saved.     October  brought  the  breakdown  of  the  Schultheiss-Patzenhofer  concern,  which, 
though  a  brewery,  had  entangled  itself  in  enterprises  in  the  fields  of  cement  and  textiles. 

3  Also  the  Schroederbank  suspended  payment  on  July  20.     Otherwise,  there  were  no 
major  bank  failures,  though  of  course  it  is  impossible  to  say  what  would  have  happened 
without  the  "bank  holidays"  and  the  subsequent  restrictions. 


932  BUSINESS  CYCLES 

That  business  transactions  were  for  a  time  not  only  hazardous  and 
unprofitable  but  in  many  cases  impossible,  stands  to  reason.  All  the  more 
important  is  it  to  stress  the  exceptional  and  extra-economic  character  of 
the  circumstances  which  caused  the  monetary  mechanism  to  produce  such 
complete  disorganization.  The  case  teaches  nothing  about  the  working 
of  the  gold  standard  and  the  merits  or  demerits  of  the  classical  or  any 
other  theory  about  price  levels,  interest  rates,  and  gold  movements.  All 
that  the  political  factor  had  put  out  of  gear.1  Foreign  ^credits  had 
financed  German  business  to  an  extent  that  would,  but  for  the  political 
factor  (Chap.  XIV,  Sec.  C),  be  inexplicable  and  is  altogether  unique. 
They  left  Germany  not  under  any  of  those  stimuli  which  are  inherent  to 
the  mechanism  of  international  finance  but  in  utter  disregard  of  them. 
It  follows  that,  since  the  efflux  of  gold  and  the  liquidation  enforced 
thereby  did  not  mean  what  they  do  mean  in  the  ordinary  course  of  com- 
mercial fluctuations,  there  was  no  reason  to  react  to  them  in  the  usual 
way,  viz.,  by  restricting  credit  and  raising  the  rate  of  interest,  and  that  the 
natural  thing  to  do  was  to  try  to  replace  the  working  capital  that  was 
being  torn  out  of  the  German  organism  by  an  expansion  of  Reichsbank 
credit  and  to  leave  the  foreign  value  of  the  mark,  for  the  time  being,  to 
itself.  Closure  of  the  banks  and  many  other  embarrassments  could  also 
have  been  avoided  by  this  method,  which,  involving  nothing  but  the 
substitution  of  a  fiduciary  basis  of  credit  for  the  gold  basis  that  had  been 
destroyed  by  extra-economic  events,  would  in  no  sense  have  been 
"inflationary."  Many  observers  believed  at  the  time  and  most  of  them 
believe  now  that  the  government  and  the  Reichsbank  were  completely 
blind  to  this  and  that,  actuated  by  an  uncritical  belief  in  an  outworn 
fetish  and  oblivious  of  anything  else,  they  clung  to  the  gold  standard  and 
precipitated  and  intensified  the  catastrophe  by  punitive  measures.  It  is 
suggested  that  this  was  not  so. 

True  enough,  government  and  Reichsbank  were  in  no  mind  and  in  no 
position  to  declare  for,  and  to  act  consistently  on,  an  "expansionist" 
policy.  They  were  not  prepared  to  face  the  domestic  and  international 
obloquy  which,  at  a  time  when  abandoning  the  gold  standard  was  not  yet 
made  respectable  by  the  example  of  England  and  America,  would  in 
that  case  assuredly  have  burst  upon  them.  Fears  that  the  foreign  value 
of  the  mark  would  be  attacked  by  speculators  and  fall  to  lows  which 
would  then  tend  to  induce  real  inflation,  and  that  at  home,  the  fetish 
once  removed,  there  would  be  no  halt  to  really  inflationary  expansion  of 
the  circulating  medium,  were  not  unjustified.  Respect  for  treaty 

1  Given  that  fact,  it  may,  of  course,  be  argued  that  that  mechanism  cannot  be  relied 
on  and  that  a  suitable  policy  and  theory  of  precisely  that  situation  must  be  worked  out. 
This  is  what — in  a  nutshell — the  rest  of  this  paragraph  attempts  to  do.  We  are  concerned 
only  with  diagnosis.  The  writer  has  no  intention  of  arguing  for  "classical"  views. 


THE  WORLD  CRISIS  AND  AFTER  933 

obligations,  promises,  and  recent  enactments  counted  for  something. 
And,  what  perhaps  weighed  most  of  all  with  the  leading  statesman,  that 
remedy  might  have  barred  the  way  toward  fundamental  normalization,1 
which  had  been  embarked  upon  with  such  difficulty  and  was  to  preclude 
all  possibility  of  the  recurrence  of  similar  situations.  He,  therefore,  kept 
to  the  policy  inaugurated  before — the  emergency  decree  of  June  5  marks 
another  step — dealing  with  every  emergency  that  confronted  him  in  the 
hope  that  it  would  be  the  last  of  its  kind.  Nor  was  this  hope  without 
justification.  In  June  it  seemed  to  come  true:  President  Hoover's 
moratorium  plan,  though  in  its  nature  no  remedy  for  the  business  situa- 
tion of  the  moment,2  might  have  brightened  the  atmosphere  and  put  a 
stop  to  the  influence  of  the  political  component  of  the  situation.  If  it 
had,  Germany's  position  would  have  been  all  the  better  for  having 
adhered  to  financial  decorum.  It  did  not.  But  when  that  became  clear, 
the  price  exacted  by  official  adherence  to  an  orthodox  policy  had  been 
largely  paid. 

This,  however,  expresses  but  one  aspect  of  the  policy  of  those  months. 
Closer  scrutiny  shows  that  the  other,  the  expansionist  one,  was  by  no 
means  absent.  We  start  by  recalling  the  general  features  of  the  inter- 
national business  situation  down  to  the  end  of  May.  In  accordance  with 
them,  money  rates  fell  everywhere — the  Dutch  bank  rate  was  reduced  to 
2  per  cent  as  late  as  May  16.  The  unrest  among  short  balances,  which 
put  an  end  to  this,  was  originally  confined  to  Germany  (Austria  and 
Hungary  being  affected  in  sympathy).  At  first  the  Reichsbank  did  not 
react  at  all.  It  was  only  after  having  lost  about  600  millions  of  gold  and 
exchange  from  June  1  to  June  11  and  when  it  was  faced  with  panic 
demands  on  the  twelfth  and  the  thirteenth,  that  it  raised  its  rate  from 
5  to  7  per  cent.  In  doing  so  it  acted  on  the  hypothesis  that  this  would 
induce  member  banks  to  use  their  own  reserves  and  impress  at  least  some 
of  the  components  of  that  demand.  Tension,  in  fact,  relaxed  and  on  the 
strength  of  the  hope  raised  by  the  Hoover  moratorium  no  further  restric- 
tions were  resorted  to.  The  Reichsbank  notes  were,  owing  to  the  credit 
of  100  million  dollars  mentioned  above,  still  covered  to  40.1  percent  at 

1  Nobody  can  say  whether  it  would  have.     It  may  be  plausibly  argued,  especially  from 
comfortable  combination  room  armchairs,  that  a  bold  announcement  of  a  policy  of  expan- 
sion would  have  made  normalization  easier.     Given  German  conditions,  the  writer  does 
not  share  any  such  belief.     But  it  may  well  have  been  a  fifty-fifty  chance. 

2  The  plan,  proposed  on  June  20,  met  with  French  opposition  and  was  not  definitively 
adopted  until  August  11.     To  this  delay  it  has  become  a  tradition  to  ascribe  its  failure  to 
take  effect.     But  to  the  writer  the  case  seems  rather  to  illustrate  the  limitations  of  the 
"psychological"  factor.     The  plan  was  certainly  calculated  to  remove  a  cloud  from  the 
horizon.     But  to  the  individual  businessman,  who  the  writer  believes  is  less  swayed  by 
irrational  moods  than  is  commonly  believed,  it  held  out  little  promise  for  the  moment. 
This  is  the  reason  why  it  produced  nothing  but  an  ephemeral  stock  exchange  boom. 


934  BUSINESS  CYCLES 

the  end  of  the  month,  in  spite  of  the  new  wave  of  withdrawals  that  was 
incident  to  the  rumors  about  the  Darmstaedter  Bank.  With  the  crisis 
gathering  momentum  and  gold  holdings  falling  precipitously,  the  Bank 
indeed  went  to  10  (July  15)  and  15  (Aug.  1)  per  cent — from  which  it  then 
receded  on  Aug.  12  (to  10  per  cent),  on  Sept.  1  (to  8  per  cent),  and  on 
Dec.  9  (to  7  per  cent) .  But  at  the  same  time  it  did  replace  the  gold  asset 
by  bills  and  advances.  Excluding  the  bills  that  were  sent  abroad  for 
rediscount,  it  held  by  Nov.  30  over  4.2  billion  marks  of  bills,  treasury 
bills,  and  advances  (Lombard),  as  against  less  than  2  at  the  end  of  May.1 
This  was  made  possible  by  the  suspension  of  the  clause  concerning  the 
40  per  cent  cover  of  bank  notes  and  by  the  foundation  (July  26)  of  the 
Acceptance  and  Guarantee  Bank,  which  merely  served  to  create  financial 
paper  for  rediscounting.  Thus  the  authorities  went  as  far  as  was  com- 
patible with  upholding  the  foreign  value  of  the  mark — further,  in  fact, 
for  that  value  was  upheld  by  measures  that  rendered  it  even  then  largely 
nominal — and  less  pressure  was  exerted  on  business  than  is  commonly 
believed.  That  the  severe  slump  in  production  and  general  business 
which  occurred  in  the  third  quarter  of  the  year  was  primarily  due  to  that 
monetary  crisis  we  do  not  by  this  statement  call  into  question;  but  it 
was  due  to  the  whole  complex  of  causes  and  consequences  of  that  crisis 
and  not  to  central  bank  action  or  interest  rates  in  particular.  On  the 
contrary,  these  played  no  role  among  the  causes  and  but  a  minor  one 
among  the  consequences. 

Neglecting  minor  difficulties,  weakness  of  the  mark  in  foreign  markets 
among  them,  we  may  say  that  the  stage  was  set  for  some  improvement 
when  England  abandoned  the  gold  standard,  which  for  Germany  was, 
of  course,  a  serious  matter.2  The  question  why  Germany  did  not  follow 
suit  like  the  Northern  countries  has  never  been  answered  quite  satis- 
factorily.3 Perhaps  a  partial  answer  may  be  found  in  the  consideration 
that  some  of  the  effects  of  devaluation  can  also  be  secured  by  scaling 

1  Liabilities  of  the  six  big  Berlin  banks  decreased  by  about  3  billions  during  the  same 
period. 

2  In  some  industries  the  flow  of  orders  from  England  suddenly  stopped  and  many 
cancellations  were  received.     Immediate  effects  were,  nevertheless,  not  great.     Ulterior 
effects  are  difficult  to  estimate.     We  will  pass  by  the  stock  exchange  and  banking  events  of 
September,  which  also  were  among  the  consequences  and  eventually  led  to  Germany's 
application  for  the  convocation  of  the  Young  Plan  Committee.     Owing  to  help  extended, 
especially  from  the  Federal  Reserve  Bank  of  New  York,  the  Reichsbank  weathered  the 
storm  comparatively  well.     At  the  end  of  October  its  notes  were  still  covered  to  nearly 
30  per  cent. 

8  The  writer  understands  that  it  has  been  answered,  by  a  man  who  ought  to  know,  to 
the  effect  that  England  "would  not  consent."  But  what  could  she  have  done?  Nor 
would  it  be  satisfactory  to  repeat  the  string  of  arguments  glanced  at  above.  For  England's 
example  made  all  the  difference,  and  linking  the  mark  to  the  paper  pound  would  have 
staved  off  the  danger  of  disorganization  through  inflationary  fears. 


THE  WORLD  CRISIS  AND  AFTER  935 

down  all  prices  and  incomes  (see  below,  Sec.  E).  It  might,  hence,  have 
been  argued  that  government  would  only  have  to  go  on  with  its  policy 
in  order  to  achieve  a  similar  end  by  another  route.  This  is  in  any  case 
what  they  attempted  to  do  by  various  measures,  among  which  the 
emergency  decree  of  Dec.  8,  reducing  wages  and  interest  charges,1  was 
the  most  important.  Whatever  may  be  thought  about  this — the  writer 
is  far  from  wishing  to  defend  it — it  was  not  simply  nonsense  nor,  as  can 
be  easily  inferred  from  the  similarity  with  devaluation,  was  it  deflation 
in  any  ordinary  sense.  And  no  dire  consequences  followed.  Production 
during  the  first  quarter  of  1932  was  indeed  at  a  much  lower  level  than 
during  the  first  quarter  of  1931 — although  by  less  than  the  first  quarter 
of  1931  was  below  that  of  1930,  in  spite  of  the  greatly  increased  impedi- 
ments of  international  trade,  which  reduced  value  of  total  exports  to 
little  more  than  half  of  the  1929  maximum — but  decline  had  come  to  a 
stop,  the  Berlin  Institute  figures  corrected  for  seasonal  being  for  those 
3  months  (1928  =  100):  52.9,  55.1,  56.7  (and  a  little  higher  for  the  second 
quarter).  There  was  a  relapse  to  a  new  low  in  July  and  August,  after 
which  the  upswing  definitively  set  in. 

These  figures  have  been  mentioned  because  they  bear  upon  the  remain- 
ing question,  whether  that  upswing  can  have  been  due  to  the  different 
policy  that  was  adopted  in  the  second  half  of  the  year.  The  negative 
answer  they  clearly  suggest — there  is  no  earthly  reason  to  believe  that 
the  movement  along  the  bottom  in  the  first  two  quarters  should,  barring 
accident,  have  been  followed  by  anything  else  but  recovery — is  supported 
by  a  survey  of  what  was  actually  done.  Some  relief  was  given  to  the 
taxpayer  by  the  not  uninteresting  measures  embodied  in  what  was  termed 
the  Papen  plan,2  and  more  important  than  the  actual  relief  afforded  was 
the  spirit  of  this  policy,  which  aimed  at  nursing  the  system  instead  of 
harassing  it  incessantly.  A  public  works  program  was  also  inaugurated, 
subsidies — not  altogether  absent,  however,  under  the  preceding  govern- 
ment— were  given  for  various  purposes,  especially  for  the  repairing  of 
houses,  and  industrial  construction  was  encouraged.  But  quantita- 

1  The  decree  of  Nov.  17  had  already  extended  help  to  agricultural  debtors,  the  situation 
of  whom  was  indeed  becoming  desperate.     This  decree  may  be  characterized,  as  may  some 
measures  later  on  taken  elsewhere,  as  an  attempt  at  regulated  bankruptcy,  again  the  same 
thing  which  devaluation  aims  at  by  another  technique.     This  measure  and  the  anticipa- 
tion of  similar  ones  intensified  the  severe  fall  in  the  price  of  German  bonds,  whether  payable 
in  marks  or  in  a  foreign  currency,  which  occurred  both  at  home  and  abroad  in  the  last  2 
months  of  the  year.     In  New  York  German  6  and  6%  per  cent  bonds  fell  below  20  per  cent 
by  Dec.  4.     But  the  incessant  stream  of  restrictions  also  contributed  toward  making 
balances  in  and  claims  on  Germany  all  but  valueless  to  the  foreigner. 

2  Its  most  interesting  feature  consisted  in  the  issue  of  certificates  for  tax  payments 
which,  as  it  were,  embodied  future  tax  remissions  and  were  made  eligible  for  discount  at 
member  banks  and  the  Reichsbank  (Steuergutscheine) . 


936  BUSINESS  CYCLES 

lively  all  that  was  palpably  inadequate  to  produce  a  contracyclical  move- 
ment or  even  to  lead  out  of  depression.  No  observer  of  contemporaneous 
developments  has  ever  claimed  that  such  homoeopathic  treatment  could 
do  that.  Ample  justice  is  done  to  that  policy  if  the  steadiness  of  the 
subsequent  recovery  is,  in  part  at  least,  recorded  to  its  credit.  To  this 
it  seems  in  fact  to  be  entitled  precisely  on  account  of  its  moderation, 
which  prevented  hectic  spurts  with  consequent  relapses.  But  it  would 
be  absurd  to  attribute  the  turn  of  the  tide  either  to  the  mechajiical  effect 
of  the  few  dozens  of  millions  by  which  expenditure  can  at  best  have  been 
increased  before  the  definitive  upturn  occurred,  or  the  anticipations  which 
may  have  been  induced.1  There  was,  it  is  true,  the  Lausanne  agreement 
(signed  July  9,  1932),  which  reduced  the  German  liability  on  reparation 
account  to  714  million  dollars.  But  apart  from  the  fact  that  we  are  apt 
to  exaggerate  the  effects  on  short-run  business  behavior  of  events  that 
do  not  immediately  alter  the  data  with  which  the  individual  firm  works, 
it  is  wrong  psychology  to  argue  that  Germans  must  have  felt  relieved. 
By  that  time  most  of  them  resented  the  idea  of  any  further  payments, 
and  this  attitude  had  been  strengthened  by  the  international  committee's 
report  (Dec.  24,  1931),  which  had  stated  Germany's  complete  inability 
to  pay  reparations.  The  effects  of  the  Lausanne  agreement  were,  hence, 
to  say  the  least,  doubtful  and  there  can  in  any  case  be  no  question  of  their 
having  more  than  counterbalanced  the  really  and  directly  depressing 
effect  that  emanated  from  the  steadily  thickening  fog  of  hostile  tariffs  and 
quota.  Comparative  normalization  of  monetary  conditions  was,  of 
course,  more  important.  The  Reichsbank  rate  went  down  to  4  per  cent 
in  the  course  of  1932.  But  this  cannot  be  listed  among  external  factors, 
since  it  was  the  natural  concomitant  of  the  passing  of  deep  depression 
and,  if  anything,  overdue. 

3.  Incidents,  Accidents,  and  Policy  in  the  United  States. — In  this 
country  the  principles  of  a  recovery  policy  began  to  take  shape  as  soon 
as  the  plateau  on  which  business  moved  during  the  first  half  of  1931 
gave  way.  This  new  dive  in  the  third  quarter  of  1931,  which  was  deeper 
than  we  should  have  expected,  was  but  to  a  minor  extent  due  to  the 
unpleasant  experiences  this  country  had  had  in  the  role  of  a  creditor 
nation.  Losses  from  defaults  of  foreign  debtors  on  long-term  loans  and 
from  the  depreciation  or  worse  of  other  long-term  investments,  however 
considerable,  were  small  relatively  to  the  size  of  the  organism.2  More 

1  The  date  of  the  decree  which  put  the  Papen  plan  into  force  is  Sept.  5.     If  the  Sep- 
tember figure  of  the  index  of  production  was  due  to  that  modest  though  well-conceived 
measure,  Mr.  von  Papen  has  indeed  wrought  a  miracle. 

2  But  such  considerations  as  that  the  value  of  shares  of  foreign  concerns  was  only  1.5 
per  cent  of  the  total  value  of  shares  traded  on  the  New  York  stock  exchange  and  that  the 
former  only  participated  to  about  the  same  extent  in  the  total  depreciation,  which  was 


THE  WORLD  CRISIS  AND  AFTER  987 

important  embarrassments  followed  from  the  freezing  or  worse  of  short 
credits  to  foreign  banks,  because  this  paralyzed  the  first  line  of  defense 
against  the  withdrawals  of  foreign  balances.  But  until  September  1931 
there  was  hardly  any  sign  that  serious  difficulties  would  arise  on  that  score. 
Repatriation  of  part  of  those  balances  was  to  be  expected,  because  there 
was  little  to  do  for  them  here.  Throughout  the  summer  this  was,  how- 
ever, more  than  compensated  by  movements  in  the  opposite  direction  and 
by  the  middle  of  September  the  figure  for  total  monetary  gold  stock 
stood  at  a  maximum  of  5,015  million  dollars:  the  golden  armor  of  the 
United  States  currency  seemed  to  grow  stronger  as  the  Federal  Reserve 
System  serenely  sailed  through  those  troubled  months. 

For  the  withdrawals  which  began  on  Sept.  20,  the  day  on  which  the 
Bank  of  England  suspended  gold  payments,  the  writer  has  no  other 
explanation  to  offer  but  that  this  event  suddenly  convinced  the  world  that 
no  currency  was  to  be  trusted  and  that  it  thereupon  discovered  weak- 
nesses and  inflationary  possibilities  in  the  American  position,  among  other 
things  the  possibility  of  a  domestic  run1  and  the  comparatively  small 
amount  of  gold  that  was  "free"  within  the  American  definition.  With 
this  the  fact  accords  well  that  it  was  in  the  first  instance  the  European 
central  banks  which  made  haste  to  convert  their  holdings  of  American 
exchange  into  gold.  The  gold  exports  to  France,  Belgium,  the  Nether- 
lands, and  Switzerland  which  ensued — the  reserve  banks  lost  722  million 
dollars  (exports  and  earmarks)  to  Oct.  22,  not  in  itself  a  very  serious 
matter — had  not  under  these  circumstances  the  classical  but  exactly  the 
opposite  effect:  there  was  a  "crisis  of  the  dollar"  during  which  forward 
dollars  in  Paris  fell  to  a  discount  of  from  5  to  10  per  cent,  President 
Hoover's  Second  Plan  (Oct.  6)  becoming  another  argument,  although  per- 
haps no  reason,  for  feeling  pessimistic  about  the  dollar.  But  on  Oct.  22 
Mr.  Laval  arrived  in  New  York.  And  the  fact  that  his  and  Mr.  Hoover's 
professio  fidei  sufficed  to  stop  the  efflux  and  even  to  reverse  it  proves  the 
comparative  innocuousness  of  the  intermezzo.  Moreover,  French  private 
balances  had  by  then  been  to  a  great  extent  repatriated,  and  the  Banque 
de  France  consented  for  the  time  being  to  discontinue  withdrawals  on 
condition  ( ?)  that  she  would  be  spared  a  repetition  of  her  English  experi- 
ence. Finally,  the  shock  troops  of  international  speculation  garrisoned 

roughly  58.5  billions  as  between  September  1929  and  December  1981,  of  course  tend  to 
understate  the  real  extent  of  the  relative  importance  of  total  losses  on  foreign  industrial 
investments. 

1  So  far  as  gold  is  concerned,  this  possibility  did  not  materialize  to  a  significant  extent 
at  any  time  during  those  2  years.  Total  gold  and  gold  certificates  in  circulation  kept, 
while  displaying  a  bulge  at  the  end  of  1930  and  the  beginning  of  1931,  below  the  level  of 
1927,  though  apprehensions  were  entertained  early  in  1932.  It  was  different  with  "  money" 
in  circulation.  In  this  item  the  big  increase  occurred  in  the  second  hah4  of  1931. 


938  BUSINESS  CYCLES 

between  Amsterdam  and  Zurich — 500  million  dollars  is  perhaps  an 
exaggerated  estimate,  though  it  comes  from  good  authority — had  for 
the  time  being  a  more  interesting  object  to  attack.  In  (February  and) 
May  and  June  1932  a  similar  wave  of  withdrawals  occurred,  though  on 
a  smaller  scale.  Reversal  in  the  second  half  of  the  year  this  time  left 
the  country's  total  monetary  gold  stock  slightly  increased — release  from 
earmark  and  domestic  production,  etc.,  more  than  balancing  exports — 
while  it  had  decreased,  though  only  by  133.4  million  dollars  in  1931. J 
This  practically  disposed  of  the  rest  of  French  claims. 

With  due  respect  to  "psychology,"  no  major  feature  of  the  conditions 
in  this  country  can  possibly  be  ascribed  to  either  those  gold  movements 
themselves  or  such  effect  as  they  had  (see  below)  on  the  structure  of 
interest  rates.  But  another  element  asserted  itself  more  and  more  as 
1931  wore  on,  viz.,  the  debt  and  especially  the  mortgage  situation.  Its 
importance  is  not  adequately  measured  by  the  number  of  bank  suspen- 
sions, although  it  was  formidable:  in  1931  2,298  banks  had  to  pull  down 
their  shutters  or,  more  significantly,  1,702  from  September  1931  to 
January  1932,  the  period  of  the  second  epidemic,2  which  was,  of  course,  a 
major  factor  in  the  general  slump  of  the  first  half  of  1932  and  the  then 
soaring  figure  of  business  failures.  Nor  is  it  adequately  measured  by 
the  losses  incurred  by  banks  and  other  creditors  on  bad  debts.  Certain 
classes  of  the  latter,  debts  from  installment  sales  for  instance,  behaved 
remarkably  well.  But  the  strain  and  drain  of  the  repayments  that  were 
successfully  made — Professor  Fisher's  debt-deflation — and  the  general 
awareness  of  the  fact  that  the  value  of  collateral  was  impaired  and  the 
net  worth  of  so  many  people  negative,  partly  enforced  and  partly  sug- 
gested restriction  of  operations  all  round,  pressed  on  prices,  decreased 
employment.  There  is  nothing  astonishing  in  the  fact  that  this  situation 
did  not  assert  itself  fully  until  about  a  year  after  the  setting  in  of  depres- 
sion in  our  sense.  It  takes  time  (and  an  altogether  abnormal  burden  of 
debt  incurred)  for  it  to  develop  and  for  people  to  realize  it  and  to  cease 
to  act  on  the  hope  of  speedy  recovery.  And  it  will  then — this  is  not  the 
only  case  of  its  kind  as  an  analysis  of  1875—1876  would  show — cause  dents 
and  slumps  in  the  interval  in  which  deep  depression  should  be  giving  way 
to  more  gentle  descent. 

This  being  the  diagnosis  of  the  nature  of  American  "incidents  and 
accidents,"  how  far  were  their  effects  plus  the  fundamental  processes  of 
depression  influenced  by  recovery  policy  or  anything  else  government  and 
federal  reserve  banks  did  ?  That  policy  must,  again,  be  defined  in  terms 
of  actual  measures  and  not  in  ternus  of  disconnected,  often — really  as 

1  Figures  from  Federal  Reserve  Bulletin  or  statements.     The  minimum  in  the  total  of 
monetary  gold  occurs  in  June  1982,  when  it  stood  at  8919  million  dollars. 

2  From  February  to  October  1982  death  rate  was  comparatively  low. 


THE  WORLD  CRISIS  AND  AFTER  939 

well  as  apparently — contradictory,  always  inadequate  pronouncements. 
What  strikes  us  first  is  the  handling  of  the  international  situation  by 
the  coordinated  efforts  of  the  government  and  the  reserve  system  with  a 
view,  (a)  to  avoiding  or  mitigating  breakdown  of  foreign  credit  structures 
by  direct  help  or  by  refraining  from  pressing  American  claims;  (6)  to  relax- 
ing tension  by  the  moratorium  of  political  payments; and  (c)  to  minimizing 
repercussions  on  the  domestic  money  market.  As  regards  the  last  point, 
the  aim  was  achieved  to  a  greater  extent  than  is  generally  believed.  We 
have  seen  that  prompt  action  had,  soon  after  the  crash  of  1929,  reduced 
money-market  conditions  to  a  state  of  ease  which — for  "natural"  rea- 
sons— prevailed  through  1930  and  still  more  so  in  the  first  part  of  1931. 
The  rate  of  the  Federal  Reserve  Bank  of  New  York  went  down  to  1.5  per 
cent  on  May  7.  Nevertheless,  the  reserve  system  bought  from  June  to 
August  another  130  millions  of  governments,  with  the  result  not  only  that 
member  banks  in  the  main  centers  accumulated  considerable  excess 
reserves  and  that  bond  yields  were  forced  down,  but  also  that  new  financ- 
ing revived.1  But  the  efflux  of  short  balances  after  Sept.  20  put  member 
banks  into  debt  again  and  also  forced  many  of  them  to  sell  bonds.  And 
until  the  Glass-Steagall  Bill  (signed  February  1932)  removed  the  obstacle 
to  their  borrowing  from  reserve  banks,  which  lay  in  the  scarcity  of  eligible 
paper,  by  permitting  loans  on  hitherto  noneligible  collateral,  and  the 
obstacle  to  further  open-market  purchases  of  the  reserve  system  by  per- 
mitting governments  to  be  used  as  collateral  for  Federal  reserve  notes, 
that  is,  through  5  months,  there  was  indeed  that  "deflationary  pressure" 
of  which  so  much  has  been  made  by  some  students  of  the  depression.  But 
this  pressure  was  altogether  unequal  to  the  inferences  that  have  been 
drawn  from  it. 

Although  the  rediscount  rate  of  the  Federal  Reserve  Bank  of  New 
York  was,  in  obedience  to  the  rules  of  financial  tradition,  raised  to  2.5 
per  cent  on  Oct.  9  and  to  3.5  on  Oct.  15  and  kept  there  to  Feb.  26,  when 
it  was  reduced  to  3  per  cent,2  rates  charged  by  member  banks  to  cus- 
tomers, which  had  reached  a  monthly  minimum  of  3.93  per  cent  for 
September  1931,  increased  to  a  monthly  maximum  of  but  4.72  for  March 
1932  (New  York  City;  Federal  Reserve  Bulletin).  This  is  the  real  test 
of  the  severity  of  the  pressure  exerted,  and  on  the  strength  of  the  most 
common  experience  it  disposes  of  the  idea  that  there  was  discouragement 

1  While  this  was  going  on,   however,   production  and  employment,  from  June  on, 
decreased  strongly.     Once  more,  such  an  experience,  while  not  in  itself  sufficient  to  dispose 
of  certain  theories  about  the  efficacy  of  either  rates  or  open-market  operations,  should  not 
be  lightly  brushed  aside.     In  this  case  it  cannot  be  urged  that  the  power  of  the  spiral  was 
unbreakable.     For  5  months  had  preceded  during  which  business  had  been  "looking  up." 

2  Buying  rate  on  acceptances  was  reduced  on  Jan.  12.     Rediscount  rate  returned  to 
2.5  per  cent  on  June  24,  in  sovereign  disregard  of  the  then  gold  outflow. 


940  BUSINESS  CYCLES 

of  industrial  and  commercial  business.  Immediately,  however,  after 
the  president's  signature  had  been  affixed  to  the  Glass-Steagall  Act,  the 
reserve  system  embarked  upon  the  biggest  open-market  operation  in  its 
history,  buying  $1.11  billion  governments  from  March  to  August,  driving 
down  rates  all  round  and  piling  up  excess  reserves,  more  than  balancing 
the  effect  of  the  gold  efflux  in  May  and  June  which  this  policy  produced. 
This,  of  course,  is  the  point  at  which  a  post  hoc  ergo  propter  hoc  argument 
is  most  likely  to  suggest  itself.  It  need  not,  however,  detain  us,  since 
the  redundance  of  the  facilities  created  is  obvious.1 

The  National  Credit  Corporation  (articles  filed  Oct.  13,  1931),  the 
Home  Loan  Banks  (bill  containing  the  Glass  rider2  signed  July  22,  1932) 
and  the  Reconstruction  Finance  Corporation  (the  original  bill  signed 
Jan.  22,  the  extension,  Emergency  Relief  and  Construction  bill,  July  21) 
represent  the  attempts  made  to  mend  what,  in  fact,  were  the  most 
important  consequences  of  domestic  "accidents  and  incidents,"  the 
removal  of  which  would  allow  the  system  to  recover.  The  inadequacy 
of  the  first  measure  to  improve  materially,  let  alone  to  save,  the  banking 
situation — or  else  to  make  the  economic  system  immune  to  it — is  obvious, 
as  are  the  limitations  of  the  second,  which,  however,  within  its  limits  did 
something  toward  improving  things  in  one  sector  of  the  mortgage  embro- 
glio.  Both  were  far  surpassed  in  importance  by  the  third,  which, 
especially  as  extended  by  the  Emergency  Relief  and  Construction  Act, 
pegged  a  number  of  tottering  structures,  thus  stopping  up  some  sources 
of  infection  from  which  cumulative  disorders  would  otherwise  have  spread, 
especially  among  banks  and  trust  companies,  railroads,  building  and 
loan  associations,  insurance  companies,  and  mortgage  loan  companies. 
By  Sept.  30,  1932,  the  grand  total  actually  advanced — not  merely 
authorized — amounted  to  nearly  1.2  billion  dollars,  of  which  185  millions 
had  then  been  already  repaid,  and  the  corporation  had  issued  750  million 
dollars  of  3.5  per  cent  notes,  600  of  which  were  taken  by  the  Treasury. 
These  few  data3  suffice  to  indicate  the  aims  and  financial  nature  of  the 
measure  during  the  first  8  months  of  the  Corporation's  life  and  to  appraise 
the  kind  and  extent  of  the  influence  it  can  have  exerted  on  the  economic 
processes  around  the  lower  turning  point  of  the  index  of  production. 
Primarily  intended  as  a  support  to  banks  and  cognate  institutions,  and 
as  an  agency  to  carry  part  of  the  burden  of  loans  that  were  noneligible 

1  This  does  not  necessarily  imply  adverse  criticism  of  that  measure.  To  create  redun- 
dant facilities  in  order  to  provide  for  any  requirements  recovery  might  entail  may  be  a 
reasonable  thing  to  do,  even  though  not  likely  to  induce  an  upward  movement  in  a  business 
community  that  does  not  use  the  funds  it  already  has  or  could  have. 

8  The  Glass  rider  conferred  an  additional  circulation  privilege  on  government  bonds 
which  extended  the  national  bank-note  issue,  and  would  have  been  mentioned  in  the 
preceding  paragraph  if  it  had  been  thought  of  sufficient  importance. 

8  They  are  from  the  Corporation's  third  report  to  Congress. 


THE  WORLD  CRISIS  AND  AFTER  941 

in  the  sense  of  the  reserve  bank  legislation,1  its  scope  naturally  included 
the  only  type  of  big  business  that  was  seriously  threatened,  railroads. 
The  rationale  of  this  is  as  obvious2  as  are  the  considerable,  if  negative, 
results:  additional  disasters  were  averted,  but  not  much  positive  impulse 
was  imparted  by  it. 

Also  something  was  done  under  this  scheme,  especially  in  its  extended 
edition,  for  agricultural  credit  institutions — new  regional  agricultural 
credit  corporations  were  created,  for  instance — for  financing  the  carrying 
and  marketing  of  farm  products  and  so  on,  and  this  worked  in  with  insti- 
tutions and  policies  previously  established  for  the  benefit  of  the  agrarian 
interest.  But  as  compared  with  the  plight  of  a  large  part  of  the  agrarian 
sector,  all  that  was  done  during  those  two  years  was  surprisingly  inade- 
quate. Since  1929  the  index  of  farm  products  at  the  farm  had  fallen  by 
over  60  per  cent.3  Total  figures  of  gross  revenue  from  agricultural  produci 
tion — which,  according  to  the  estimates  of  the  Department  of  Agriculture, 
was  about  9.4  billion  dollars  in  1930,  somewhat  less  than  7  in  1931,  and 
about  5  in  1932 — and  nation-wide  indices  of  land  values — as  of  Mar.  1, 
that  index  was  115  in  1930,  106  in  1931,  89  in  1932  (1912  to  1914  =  100, 
maximum  of  170  occurring  in  1920) — do  not  tell  the  whole  tale.  For  a 
minority,  but  a  nonnegligible  one,  net  income  must  have  been  negative, 
and  for  a  considerable  minority  net  worth  of  the  farm  must  have  been 

1  The  Emergency  Relief  and  Construction  Act  also,  amending  Sec.  13  of  the  Federal 
Reserve  Act,  authorized  Federal  reserve  banks  to  discount  paper  for  individuals,  partner- 
ships, and  corporations  unable  to  secure  "adequate"  credit  accomodation  from  member 
banks  (in  our  sense).     This  move  to  force  member  banks  into  lending  freely  was  followed 
up  in  the  banking  legislation  of  1938,  1934,  and  1935.     In  itself  but  an  approach  to  Euro- 
pean practice,  it  was  part  of  a  policy  for  which  early  American  banking  experiences  must 
have  seemed  ideal,  and  a  concession  to  the  theory — which  is  fundamentally  wrong — that 
banks  hold  a  key  position  at  the  beginning  of  revival  and  that  if  their  loans  do  not  expand 
this  can  only  be  due  to  their  aversion  to  lending.     But  since  the  new  powers  were  so  very 
soberly  handled  by  the  reserve  banks,  there  is  no  need  for  us  to  go  into  this  matter. 

2  It  was  not  so,  however,  for  the  man  in  the  street,  who  did  not  see  any  relation  between 
his  own  fate  and  what  he  took  to  be  just  a  contrivance  by  which  a  capitalistic-minded 
government  tried  to  protect  fellow  capitalists  from  the  consequences  of  their  own  follies, 
leaving  the  suffering  masses  to  shift  for  themselves.     This  attitude,  which  asserted  itself 
very  soon  in  Congress  and  elsewhere  and  was  one  of  the  first  symptoms  of  the  anticapitalist 
storm  that  was  brewing,  was  strangely  neglected  by  the  administration. 

3  Nothing,  of  course,  can  be  concluded  from  that  fact  alone  or  even  from  it  plus  the 
fact  that  the  index  of  the  goods  bought  by  farmers  fell  only  by  something  over  30  per  cent. 
As  we  have  seen  before,  this  in  itself  may  spell  suffering  but  not  breakdown  and  must  be 
considered  in  connection  with  the  agrarian  revolution  of  the  twenties.     But  it  spells 
catastrophe  when  combined  with  the  fact  that  in  1930  "39  per  cent  of  the  owner-operated 
farms  of  the  United  States  were  encumbered  with  mortgages  averaging  40  per  cent  of  their 
value.     This  means  that  about  a  fifth  of  the  farms  of  the  United  States  in  that  year  had 
mortgages  representing  over  40  per  cent  of  their  value."     (Professor  John  D.  Black. 
The  Agricultural  Situation,  January,  1933,  Review  of  Economic  Statistics,  Feb.  15,  1933). 


942  BUSINESS  CYCLES 

zero  or  less.  Foreclosures  increased  rapidly  and  so  did  the  proportion  of 
forced  sales  that  were  due  to  tax  delinquency.1  Hence,  it  is  clear  that  the 
process  of  depression  was  in  the  agrarian  sector  allowed  to  go  on  even  in 
directions  in  which  it  would  have  been  most  easy  and  for  a  conservative 
government,  one  would  think,  most  imperative  to  stop  it. 

The  Emergency  Relief  and  Construction  Act  (Sec.  1,  Title  I)  marks  a 
new  departure  in  authorizing  expenditure  for  relief  and  work  relief,  a 
little  more  than  two  years  after  such  a  measure  was  indicated  according 
to  our  schema,  or  a  little  more  than  one  year  after  the  plateau  of  1931 
had  crumbled.  No  measurable  effects  can,  however,  have  emanated  from 
the  35.5  million  dollars  which  were  made  available  and  the  14.2  millions 
which  were  actually  spent  for  that  purpose  to  the  end  of  September  1932. 
The  mentality  which  had  such  difficulty  in  reconciling  itself  to  this — 
anything  but  novel  or  radical — course  of  action  and  the  persistence  of 
which  is  as  curious  as  the  violence  of  the  reaction  it  produced,  also  asserted 
itself  in  the  incessant  appeals  of  the  chief  executive  for  retrenchment  of 
public  expenditure  and  increase  of  taxation  (e.g.,  messages  and  pronounce- 
ments of  Dec.  1, 1931,  of  Jan.  8,  of  Mar.  8,  of  Apr.  4,  and  of  May  5, 1932.) 2 
In  some  cases  there  were  special  reasons  for  this,  e.g.,  the  weakness  of  the 
dollar  exchange  in  March  and  at  the  beginning  of  May  1932.  But  in  all 
cases  account  must  be  taken  of  the  fact  that  that  mentality  was,  until 
it  changed  into  its  opposite,  a  datum  of  the  situation  which  it  was  hardly 
possible  to  modify  to  just  the  extent  that  would  have  seemed  rational. 
Under  these  circumstances  a  "budget  crisis"  was  no  matter  of  indiffer- 
ence. And  adherence  in  principle  to  what  were  considered  sound  methods 
of  finance  was  not  unlikely  to  help  recovery  as  well  as  to  facilitate  fiscal 
normalization  after  the  depression,  provided  it  went  not  beyond  what  was 
necessary  to  convince  everybody  that  the  budget  would  automatically  be 
balanced  in  future  while  for  the  time  being  expenditure  was  allowed  to 
unbalance  it.  This  is  precisely  what  the  administration  actually  tried 
to  do.  Owing  to  the  insuperable  prejudice  that  defeated  the  sales  tax 
(Mar.  24), 3  little  came  of  the  first  part  of  the  program,  and  the  tax  bill 
signed  June  6  and  the  omnibus  economy  bill  signed  June  30,  1932,  cannot 
have  had  any  but  reassuring  "psychological"  effects.  But  the  second 
part  was  all  the  more  fully  carried  into  effect.  According  to  an  estimate 
used  before,  the  net4  Federal  income-generating  expenditure  amounted 

1  Professor  Black,  op.  cit.,  p.  10,  puts  that  proportion  as  high  as  one-third. 

2  It  also  asserted  itself  in  the  veto  of  the  Garner- Wagner  Relief  Bill,  which  the  writer 
does  not  find  easy  to  understand. 

8  Without  any  undesired  effects  either  during  or  after  depression,  2  billion  dollars  could 
have  been  raised  by  it  with  which  to  service  and  extinguish  a  depression  expenditure  of 
from  10  to  20  billions. 

4  Arthur  D.  Gayer,  op.  cit.,  p.  891. 


THE  WORLD  CRISIS  AND  AFTER  948 

to  1,748  million  in  1931  and  1,646  million  in  193&  (calendar  years). 
There  cannot  be  any  doubt  that  this  was  the  most  directly  effective  part 
of  the  government's  policy — the  real  Emergency  Relief — which  only 
gained  in  effectiveness  by  being  coupled  with  official  emphasis  on  those 
"  sound "  principles  that  at  first  sight  appear  to  be  at  variance  with  it. 
The  inference  is  that  it  prevented  much  potential  disaster.  Yet  since 
that  expenditure — to  which,  of  course,  a  very  low  multiplier  would  have 
to  be  applied — did  not  stop  the  shrinkage  in  total  outside  debits,  which 
fell  throughout  the  year,  the  inference  seems  reasonable  that,  although 
by  partly  compensating  the  influence  of  Incidents  and  Accidents  it 
facilitated  the  turning  of  the  tide,  it  did  not  turn  it. 

We  will  finally  glance  at  the  third  epidemic  among  banks,  which 
belongs — as  a  belated  installment — in  the  nexus  of  events  we  are  now 
surveying,  although  it  ran  its  course  entirely  within  incipient  recovery, 
lending  it  for  a  month  or  two  all  the  colors  of  deep  depression.  It  started 
in  November — the  banking  holiday  in  Nevada  declared  on  the  first  of 
that  month  may  be  taken  as  the  starting  point — gathered  momentum  in 
January  and  February  and  was  cut  short  by  emergency  legislation  on 
Mar.  9,  1933.  The  suspensions  and  holidays  spread  from  Feb.  14  on 
(Michigan;  the  first  states  to  follow  were  Indiana,  Maryland,  Arkansas, 
and  Ohio)  until  on  Mar.  5,  almost  complete  stoppage  of  banking  having 
greeted  the  new  president  on  his  inauguration,  they  had  to  be  ratified 
by  Congress.  This  time  agricultural  distress,  more  precisely  the  agri- 
cultural mortgage  situation,  was  not  merely  a  contributory  cause  but  the 
main  one,  as  is  seen  from  the  fact  that  the  hurricane  started  in  the  agrarian 
states  of  the  middle  and  farther  West  and  then  moved  to  the  East, 
thus  collecting  the  fine  for  the  neglect  of  the  agrarian  plight.  This 
suffices  to  insert  that  panic  into  our  picture.  Its  features  are  well  known. 
Distrust  in  banks,  to  some  extent  coupled  with  a  distrust  in  the  currency, 
led  to  indiscriminate  withdrawals  of  deposits  and  forced  the  banks  in 
turn  to  withdraw  currency  from  reserve  banks — member  banks  (in  the 
official  sense)  withdrew  over  1.7  billions  between  Feb.  8  and  Mar.  3 — 
and  from  New  York  correspondents  who  lost  almost  800  millions  in 
this  way.  On  Mar.  2  and  3  alone,  money  "in  circulation"  increased 
by  nearly  700  millions,  federal  reserve  credit  outstanding  by  nearly 
730.  Loss  of  reserves  and  increase  of  notes  outstanding  reduced  excess 
gold  reserves  of  the  reserve  banks  by  1.1  billion  dollars  to  400  millions. 
Domestic  difficulties  were  increased  by  a  simultaneous  efflux  of  gold 
from  the  country  which — partly  though  not  wholly  induced  by  them — 
amounted  to  over  270  millions  in  February  and  March.  The  New 
York  Reserve  Bank  had  to  rediscount  with  and  to  sell  governments 
to  other  reserve  banks.  After  the  banking  holiday  and  under  the  pressure 
of  the  Emergency  Banking  Act  of  Mar.  9  (amended  by  the  Act  of  Mar.  24) 


944  BUSINESS  CYCLES 

gold  coins  and  certificates  speedily  flowed  back,  over  600  *  millions  return- 
ing to  the  reserve  banks  before  the  end  of  March,  so  that  their  excess 
gold  reserves  were,  thereby  and  by  the  reduction  of  the  amount  required 
to  be  held  against  notes  outstanding,  increased  to  1,172  millions.  In  spite 
of  the  restrictions  to  which  gold  movements  had  been  subjected,  the  inter- 
national position  of  the  dollar  was  remarkably  strong  at  the  erid  of  March. 
Member  banks  (in  the  official  sense)  holding  about  90  per  cent  of  all 
member  bank  deposits  were  reopened  by  license  on  Mar.  15.  By  the 
middle  of  1933  the  number  of  All  Banks  (including  private  banks  under 
state  supervision  and  mutual  and  stock  saving  banks)  operating  under 
license — no  doubt  many  very  weak  ones  among  them — was  14,530.2 

The  immediate  consequences  of  that  panic,  the  new  spiral  it  set  in 
motion,  do  not  call  for  additional  comment;  but  its  ulterior  consequences 
cannot  be  too  strongly  emphasized.  It  completely  demoralized  all  classes 
and,  by  doing  so,  fundamentally  changed  the  problem  before  the  incoming 
administration.  Without  it — and  it  was  certainly  an  avoidable  incident 
— recovery  policy  would  have  been  confronted  with  an  entirely  different 
situation.  As  it  was,  the  psychic  framework  of  society,  which  till  then 
had  borne  up  well,  was  at  last  giving  way.  Nobody  for  the  time  being 
foresaw  anything  but  continuing  disaster,  and  everybody  was  resolved  not 
to  put  up  with  it  any  longer.  The  talk  about  impending  revolution 
presumably  was  nonsense;  but  it  characterizes  well  the  prevailing  state 
of  the  public  mind,  which,  bewildered  and  exasperated  to  the  utmost, 
clamored  for  political  action  in  redress  of  what  every  group  in  its  own 
way  felt  to  be  some  grievous  wrong.  Politicians  and  "intellectuals," 
suddenly  moved  into  a  position  of  saviors  and  judges,  had  a  rich  keyboard 
to  play  on.  But  the  mentality  of  the  country,  the  traditions  of  the 
victorious  party,  the  nature  of  the  catastrophe  that  had  to  be  dealt  with 
immediately,  and  the  strength  of  the  inflationist  interests  united  the 
majority  of  them  on  monetary  expansion. 

4.  Deferring  further  discussion  of  the  English  case,  we  now  return 
to  the  question  whether  other  American  and  German  time  series  confirm 
the  location  of  the  bottom  of  the  fourth  Juglar  which  above  has  been 
determined  mainly  from  the  behavior  of  physical  output.3  That  they 
otherwise  behaved  as  we  should  expect,  taking  account  of  the  incidents 
and  accidents  just  discussed,  is  obvious.  Money  rates  in  particular  do 
not  seem  to  require  additional  comment — in  Germany  the  prime  bank 

1  The  total  for  all  types  of  currency  was  $1,185  million,  most  of  which  went  towards 
canceling  reserve  credit — member  banks'  indebtedness  decreased  to  $545  million  by  Mar. 
29,  when  New  York  City  banks  again  held  excess  reserves. 

2  The  total  for  the  middle  of  1928  was  25,941  and  for  the  middle  of  1932,  18,794. 

8  It  should  be  noticed  that  industrial  consumption  of  electric  power  would  do  equally 
well,  but  not  total  power  production. 


THE  WORLD  CRISIS  AND  AFTER  945 

acceptance  rate  (Privatdiscont)  was  at  3.88  per  cent  in  December  1933, 
having  steadily  fallen  from  the  peak  in  the  second  half  of  1932  (7.95 
for  that  half  year),  and  in  this  country  the  third  banking  epidemic  failed 
to  produce  panic  rates  and  only  interrupted  the  downward  course  for  a 
short  time,  as  the  gold  panic  of  September-October  1931  had  done: 
bankers'  acceptance  rate  for  90  days'  unindorsed  bills  was  1^  per  cent 
on  Feb.  28  and  only  2  per  cent  on  Mar.  31,  1933,  other  rates  moving 
correspondingly. 

The  Harvard  A-curve  (speculation :  index  of  prices  of  all  listed  stocks) 
clearly  indicates  the  trough  for  June- July  1932  and  reacts  well  to  the 
incipient  recovery  throughout  the  third  quarter  of  that  year.  Equally 
corroborative  is  the  behavior  of  German  stock  prices,  the  index  of  which 
increased  considerably  in  the  second  half  of  the  year,  climbing  up  to  a 
little  over  half  of  its  value  for  1925.  But  outside  bank  debits  (and  dollar 
volume  of  department-store  sales)  continued  their  downward  course, 
with  but  an  insignificant  upward  movement  in  the  last  (department-store 
sales  in  the  third)  quarter,  right  into  1933,  and  thus  at  first  sight  seem  to 
cast  doubt  on  our  location  of  the  trough.1  Considering,  however,  the 
persistence  of  the  fall  in  price  level  and  the  fact  that  bank  debits  were,  of 
course,  particularly  sensitive  to  the  banking  calamity,  this  does  not  mean 
much.  Moreover,  incipient  recovery  is,  as  we  know,  compatible  with 
some  further  shrinkage  of  total  dollar  volume  of  business  operations.  It 
would  be  compatible  even  with  some  further  increase  in  failures — 
although  in  this  case  maximum  of  failures  actually  occurred  at  the  trough. 
These  and  other  symptoms2  may  be  likened  to  those  symptoms  of  disease 
which  often  show  most  markedly  in  convalescence.  The  real  question 
arises  with  respect  to  employment  and  prices. 

In  Germany  the  number  of  employed  as  per  sickness-insurance 
statistics  was  (monthly  average)  17.6  millions  in  1929,  16.3  in  1930, 
14.25  in  1931,  and  a  little  below  12.2  in  the  first  half  of  1932.  Then  it 
increased  to  October  (12.9),  to  fall  more  than  seasonally  in  December  and 
January — which,  it  will  be  recalled,  is  not  contrary  to  expectation  from 
the  experimental  schema — before  increase  had  gathered  momentum. 
The  number  of  (statistically  visible)  unemployed  rose  by  about  50  per 
cent  in  the  average  of  1931,  as  compared  with  that  of  1930,  which  was 

1  Value  of  construction  contracts  (Dodge)  increased,  however,  nonseasonally  in  the 
third  quarter,  though  they  declined  more  than  seasonally  in  the  fourth.     This  is  also  true 
of  privately  financed  (mainly  residential)  construction.     See  J.  B.  Hubbard,  The  Con- 
struction Industry  in  Depression,  Harvard  Business  Review,  January  1933. 

2  In  Germany,  for  example,  household  consumption  even  in  physical  terms  definitely 
declined  in  the  third  quarter  of  1932,  and  the  simultaneous  increase  in  the  output  of 
consumption-good  industries  at  first  went  toward  replenishing  the  stocks  of  wholesale  and, 
to  a  lesser  extent,  retail  trade.     As  to  monetary  terms,  turnover  of  cooperative  stores  per 
member  declined  in  the  third  quarter  by  3.04  marks. 


946  BUSINESS  CYCLES 

3.1  millions.  The  first  quarter  of  1932  displays  the  maximum  of  over 
6  millions,  the  figure  for  the  second  quarter  is  5.66  millions,  and  for  the 
third  a  little  over  5.2.  As  noticed  before  and  as  we  should  expect,  the 
number  of  workmen  employed  throughout  fell  less  and  increased1  less 
than  production,  and  also  lagged  behind  it.  This  is  equally  true  of  the 
United  States,  where  the  behavior  of  employment  was  very  similar. 
The  annual  minimum,  of  course,  occurs  for  1932.  More  important 
than  this,  employment  in  manufacturing  industries  begaij  to  increase, 
slightly  at  first,  in  the  second  half  of  July.  The  Federal  Reserve  Board's 
index  then  records  some  more  than  seasonal  net  increase  for  August, 
though  there  was  decrease  in  the  automobile  and  allied,  as  well  as  in 
the  machinery,  industries.  Increase  spread  in  September  (the  index 
adjusted  for  seasonal  then  showed  60.3  per  cent  of  the  1923  to  1925 
average  as  compared  with  58.8  per  cent  in  August)  and  persisted  to  the 
middle  of  November — when  employment  in  the  automobile  industry 
increased  considerably — after  which  there  was  more  than  seasonal  decline 
in  December  and  January.  Unemployment  (A.  F.  of  L.  estimate) 
behaved  accordingly.  As  in  Germany,  however,  a  new  low  point  occurred 
in  the  first  quarter  of  1933,  which  in  this  case  is  amply  accounted  for  by 
the  banking  crisis. 

But  price  levels  declined  unequivocally  through  1932  and  for  some 
time  after  in  all  countries  that  remained  on  the  gold  standard  and  in 
some  that  did  not,2  and  the  failure  of  other  series  to  display  trough 
and  recovery  in  1932  is  primarily  due  to  this  fact.  Since  this  might 
prove  disturbing  to  the  reader  who  has  followed  so  far  but  still  holds  on 
to  his  habit  of  associating  cyclical  phases  primarily  and  even  causally 
with  movements  of  the  price  level,  it  is  necessary  to  remind  him  that 
the  processes  of  recovery  do  not  require  that  price  level  should  first  rise 
or  even  cease  to  fall.  It  will  be  useful  to  recall  how  we  should  expect  it 
to  behave  at  such  a  juncture,  i.e.,  at  the  beginning  of  the  recovery  of  a 
Juglar  (preceded  by  a  Kitchin  prosperity)  which  lies  within  a  Kondratieff 
depression.  On  the  one  hand,  although  there  may  be  belated  price 

1  It  will  be  recalled  that  this  is  due  not  only  to  statistical  causes  and  to  the  fact  that 
workmen  are  not  promptly  discharged,  so  that  at  the  beginning  production  increases 
simply  in  function  of  decreasing  short-time  or  underutilization,  but  also  to  the  changes 
in  production  functions  (rationalizations). 

2  Many  individual  prices,  most  of  the  world's  staples  among  them,  do,  however,  display 
either  minima  and  recovery  or  that  flattening  out  (amidst  "hesitations")  of  curves  which 
is  the  characteristic  shape  at  the  bottom  of  a  cycle  (as  did  also  quantities).     This  may 
be  observed,  for  example,  in  the  United  States  quotations  (futures  or  spot  price)  of  cotton, 
rubber  (which  at  its  low  in  June  was  little  more  than  5  per  cent  of  its  annual  average  for 
1918),  zinc,  lead,  copper,  iron,  and  scrap  steel.     Most  of  them  lost  part  of  their  gain  in 
the  last  quarter — copper,  in  particular,  because  of  the  difficulties  in  the  cartel.     Even 
the  wool  and  petroleum  curves  flattened  out. 


THE  WORLD  CRISIS  AND  AFTER  947 

reductions — in  some  cases  induced  precisely  by  producers'  realizing  that, 
paralysis  being  over,  reductions  may  now  have  some  effect  in  stimulating 
demand1 — the  bulk  of  wholesale  prices  will,  in  fact,  recover  from  panic 
lows.  But,  on  the  other  hand,  such  "correctional"  movements  are 
superimposed  on  a  fundamental  tendency  that  works  against  their 
effect  on  the  index.  We  know  that,  and  why,  the  price  level  should  in 
every  neighborhood  of  equilibrium  be  at  a  lower  figure  than  in  the 
preceding  neighborhood,  and  in  a  Juglar  within  a  Kondratieff  depression 
this  tendency  may  result  in  its  recovery  phase  ending  up  with  a  price 
level  below  that  of  the  lower  turning  point.  Owing  to  the  violence 
of  the  break  in  prices  during  the  preceding  Juglar  depression,  this  would 
not  have  been  likely  to  happen  in  this  case  even  without  the  subse- 
quent efforts  to  raise  prices  by  political  action.  But  that  any  rise 
in  prices  that  may  have  been  due  in  reaction  to  depressive  excesses  was 
slow  in  coming  about  and  even  that  the  fall  in  wholesale  prices  arid  in  cost 
of  living  persisted  for  several  months  after  the  turning  point  of  the  cyclical 
process  is  neither  surprising  nor  a  reason  to  question  our  dating,  let 
alone  to  date  the  cyclical  trough  February  1933.  It  should  be  observed 
that  this  argument  is  independent  of  the  fact  that  for  this  country 
the  bank  holidays  and  the  events  that  led  up  to  them  provide  a  special 
and,  according  to  our  diagnosis,  an  "accidental"  cause  for  that  trough, 
and  therefore  suffice  to  rule  out  that  dating;  for  although  this  is  true, 
the  fall  in  price-level  graphs  also  went  on  in  other  countries.2 

Particulars  of  price  movements  would  merit  discussion.  We  must 
limit  ourselves  to  a  reference  to  the  rich  literature  on  the  subject3  and  to 
the  following  remarks.  First,  barring  the  effects  of  monetary  changes, 

1  Or  also  by  producers'  so  vigorously  responding  to  an  anticipated  rise  that  instead  of 
it  a  fall  ensues. 

2  The  influence  of  the  epidemic  among  banks  is  not  exactly  measurable  but  very  visi- 
ble.    The  minimum  of  the  cost-of-living  index  (National  Industrial  Conference  Board) 
and  the  minimum  of  the  B.L.S.  index  of  wholesale  prices,  both  of  which  occur  in  February 
1933,  are  obviously  linked  up  with  it,  although  these  dents  lie  in  a  declining  interval  which 
may  be  interpreted  in  the  light  of  our  schema:  there  was  a  rally  in  prices  in  the  summer  of 
1932,  which  is  absent  or  practically  so  in  Germany — where,  however,  the  index  of  sensitive 
prices  increased  from  45.3  per  cent  of  the  1913  figure  to  53  per  cent  during  the  second  half 
of  the  year — but  is  well  marked  in  this  and  some  other  countries,  and  which  we  may  asso- 
ciate with  the  Kitchin  prosperity  phase  at  the  bottom  of  the  Juglar.     The  B.L.S.  index 
shows  a  small  rise  in  July,  and  many  leading  commodities  advanced  considerably  in  August, 
while  some  that  had  advanced  before  declined.     This  up  and  down  continued  in  September, 
for  which  the  index  showed  hardly  any  change.     By  the  end  of  September  decline  had 
become  dominant,  and  for  October  the  index  was  one  point  lower.     This  decline  continued 
during  November  and  December,  and  the  January  figure,  4.2  points  below  that  of  August, 
may  already  have  been  affected  by  the  banking  crisis.     But  the  fact  that  there  was  much 
greater  price  stability  as  compared  with  1930  and  1931  is  beyond  doubt. 

3  Mention  should  be  made,  in  particular,  of  Professor  F.  C.  Mills,  Prices  in  Recession 
and  Recovery,  1936. 


948  BUSINESS  CYCLES 

the  fall  in  price  level  and  in  cost  of  living  wa§  remarkably  uniform.  When 
all  qualifications  on  the  score  of  comparability  have  been  made,  it  is  still 
significant  to  note,  for  example,  that  the  American  and  German  indices  of 
wholesale  prices  yield,  if  1929  is  taken  as  base  year,  practically  coincident 
curves  and  that  cost  of  living  fell  from  1929  to  1932  by  about  22  per  cent 
in  both  countries.  And  this  is  not  merely  due  to  the  influence  of  the 
international  prices  which  enter  the  indices. 

Second,  it  follows  from  the  argument  of  this  book  as  a  whole  and  more 
particularly  from  what  has  been  said  in  the  preceding  chapter  that  that 
fall  is  not  adequately  characterized  by  being  called  an  unforeseen  disaster 
or  a  catastrophe  of  the  price  structure  wantonly  wrought  by  monetary 
factors  or  the  vicious  spiral — debt  deflation,  in  particular — and  the  like. 
No  doubt  these  and  other  elements  contributed  to  the  violence  and,  in 
many  individual  cases  of  raw  materials  and  semifinished  products,  to  the 
extent  of  the  drop  from,  say,  the  middle  of  1931  on,  when  wholesale  prices 
had  fallen  by  about  22  per  cent  of  the  1929  average.  But  it  has  been 
shown  that  a  price  level  markedly  below  that  of  1913  was  what  would 
certainly  have  prevailed  without  the  war  and  what,  even  with  the  war, 
was  bound  to  emerge  in  time  as  a  result  of  the  evolutionary  mechanism 
and  as  a  consequence  of  and  adaptation  to  the  industrial  revolution 
of  the  age.1  The  reader  will  think  as  he  pleases  about  the  desirability 
or  otherwise  of  letting  this  process  have  its  way.  But  he  should  not 
overlook — however  he  may  appraise — its  economic  function  and  its 
potential  long-run  results.2 

Third,  it  has  often  been  pointed  out  how  differently  different  groups 
of  commodities  were  affected  and  how  rigorously  the  price  system  was 
changed  thereby.  The  difference  between  the  behavior  of  prices  of  raw 
materials  and  the  behavior  of  prices  of  manufactured  products  has 
attracted  particular  attention  and  been  held  to  have  not  only  reflected 
but  also  intensified  the  growing  disequilibrium.  All  this  is  true  to  some 
extent,  but  it  does  not  tell  the  whole  story  either  as  to  facts  or  as  to  infer- 
ence. The  minimum  of  the  B.L.S.  index  of  wholesale  prices  (February 
1933)  was  62  per  cent  of  its  value  for  July  1929.  The  National  Bureau's 
index  of  physical  volume  of  total  production3  gives  for  the  minimum  year 

1  It  should,  however,  be  recalled  that  the  fall  in  the  wholesale  price  index  from  the 
beginning  of  1930  to  the  middle  of  1932  was  much  sharper  than  the  fall  from  the  beginning 
of  1873  to  the  middle  of  1875,  which  was  part  of  an  almost  steady  decline  that  (not  to  count 
the  drop  from  the  Civil  War  peak)  lasted  almost  uninterruptedly  from  1866  to  the  middle 
of  1879. 

8  No  one  who  is  not  a  socialist  can,  of  course,  wholly  approve  of  these  results.  For 
one  of  them  would  have  consisted  in  large  strata  of  independent  or  semi-independent  agra- 
rian, commercial,  and  industrial  business  being  weeded  out,  hence  in  a  long  step  toward 
socialism.  But  this  does  not  matter  here. 

8  See  Professor  Mills,  National  Bureau  of  Economic  Research  Bulletin,  Feb.  20,  1933, 
description  of  the  index  on  p.  6.  On  prices  see  the  Bulletin  for  Oct.  31,  1933. 


THE  WORLD  CRISIS  AND  AFTER  949 

almost  exactly  62  per  cent  of  the  annual  figure  for  1929.  But  one  com- 
ponent, construction,  contracted  to  31.5  per  cent,  while  the  index  of 
building  material  prices  was  still  75.9  per  cent,  and  the  hourly  wage  rate 
about  80  per  cent  of  their  1929  values  at  the  trough  in  February.  Here 
we  have  an  obvious  case  of  maladjustment.  On  the  other  hand,  agricul- 
tural raw  products  show  no  influence  of  the  depression  on  output  for  1931 
and  at  most  a  small  one  for  1932:  "the  farmer  accepts  the  cut"  both 
because  he  works  under  conditions  of  competition  and  because  of  the 
technological  peculiarities  of  his  production.  The  opposite  reasons  do 
not,  however,  wholly  explain  the  fact  that  output  of  mineral  raw  materials 
contracted  a  little  more  than  total  output;  for  there  were,  in  addition, 
elements  of  prime  costs  which  failed  to  fall  correspondingly,  especially 
wages.1  These  elements — selling  charges  among  them — acquire,  of 
course,  increasing  importance  for  manufacturing  industry  as  we  proceed 
toward  the  finished  article  of  consumption,  and  there  is,  hence,  little  to 
be  surprised  at  in  the  February  figure  of  the  price  index  for  processed 
nonfood  consumers'  goods  (73.2  per  cent  of  that  of  July  29).  But  many 
manufacturing  industries  also  "took  the  cut,'*  for  instance,  petroleum 
refining,  food,  tobacco,  and  leather  product  industries,  paper  and  print- 
ing, clothing,  and  house  furnishing.2  Equipment  industries  did  so  to  a 
much  lesser  extent  or  not  at  all.  But  precisely  in  their  case,  reduction 
of  prices  would  hardly  have  stimulated  demand. 

These  observations  are  in  keeping  with  the  view  previously  arrived  at 
on  the  subject  of  price  rigidities.  They  are  also  relevant  to  the  question 
of  the  nature  and  consequences  of  the  disruption  by  the  depressive  process 
of  the  preexisting  structure  of  (relative)  prices.  Our  model  does  lead 
us  to  expect  dispersions — because  of  rigidities  as  well  as  for  other  reasons 
— which  spell  disequilibrium.  But  it  does  not  follow  that  every  change 
wrought  by  depression  in  the  price  system  necessarily  falls  into  that 
category  or  that  return  to  equilibrium  necessarily  requires  the  recon- 
struction of  the  preceding  system  of  relative  prices.  The  contrary  may 
well  be  the  case — an  example  is  the  price  of  copper,  which  the  opening 
of  new  sources  of  supply  had  turned  into  an  untenable  maladjustment  that 
would  simply  be  conserved  by  uncritical  attempts  to  restore  either  price 
or  income  parities.  Nor  does  it  follow  that  every  change  in  the  price 
system  which  does  fall  into  that  category  necessarily  impedes  recovery. 
It  may  facilitate  it  or  be  harmless.  An  example  of  the  first  possibility  is 

1  The  farmer  "accepts  cuts"  in  two  different  roles:  first,  as  a  producer  and,  second,  as  a 
laborer. 

2  In  a  not  unimportant  number  of  cases  actual  price  reductions  were  much  greater  than 
would  appear  from  indices  based  upon  list  prices.     In  a  still  more  important  number  of 
cases  comparison  between  the  fall  in  the  prices  of  agricultural  and  the  fall  in  the  prices  of 
industrial  products  is  rendered  meaningless  by  the  changes  in  the  quality  of  the  latter, 
for  which  it  would  be  necessary  to  correct. 


950  BUSINESS  CYCLES 

afforded  by  any  panicky  and  temporary  drop  in  foreign-produced  raw 
materials;  an  example  of  the  second,  by  the  short-term  rigidity  of  prices 
of  equipment  goods,  for  recovery  does  not  typically  start  from  expansion 
of  real  investment.  Many  particulars  were,  however,  different  in  Ger- 
many. We  will  merely  mention  that  agrarian  prices  did  not  fall  anything 
like  as  much  as  in  this  country.  In  December  1932  they  were  about 
65  per  cent  of  the  1929  average,  keeping  about  in  step  with  the  total 
index  and  practically  avoiding  the  catastrophic  slide  of  the  American 
prices  in  1931. 

Adjusted  demand  deposits  of  reporting  member  banks  in  reserve  cities 
outside  New  York,  after  keeping  up  to  almost  the  middle  of  1931  fell 
sharply  to  May  1932,  after  which  there  was  a  small  increase  that  was 
just  about  wiped  out  in  the  first  quarter  of  1933.  "Country  banks" 
taken  separately  displayed  but  a  decreased  rate  of  decrease  in  the  second 
half  of  1932.  Net  demand  deposits  of  reporting  member  banks  outside 
New  York  City  fell  after  the  middle  of  1931,  first  at  an  increasing  then 
at  a  decreasing  rate,  the  curve  flattening  out  and  then  slightly  rising 
in  1932.  This  is  not  exactly  what  we  should  expect  but  is  accounted  for 
by  the  changes  in  the  investment  item,  which  increased  strongly  through 

1930  and  the  first  third  of  1931,  then  fell  to  the  beginning  of  1932  and 
increased  again  to  almost  the  end  of  the  year,  while  All  Other  Loans  fell 
strongly  and  almost  continuously  throughout  and  beyond.     The  index 
of  rate  of  turnover  of  demand  deposits  in  principal  cities,  which  we  owe 
to  the  Federal  Reserve  Bank  of  New  York1  declined  at  a  decreasing  rate 
near  the  end  of  1932,  when  it  began  to  rise  again. 

National  Income  produced,  evaluated  at  1929  prices,  fell  from  1930 
to  1931  by  a  larger  amount  than  it  had  fallen  from  1929  to  1930,  and  from 

1931  to  1932,  the  minimum,  by  a  larger  amount  than  from  1930  to  1931.2 
Net  corporate  income  (All  Corporations  except  tax-exempt  ones  and  life 
insurance  companies;  before  payment  of  income  tax)  became  negative  to 
the  amount  of  2,850  millions  in  1931,  and  for  1932  displays  the  maximum 
loss  of  5,200  millions.     The  number  of  corporations  reporting  loss  was 
greater  than  the  number  reporting  positive  revenue  as  early  as  1930;  in 
1931  the  relation  was  284:176;  in  1932  it  was  366:80.     Still  more  sig- 
nificant are  the  figures  of  corporate  accumulation,  though  the  limited 
value  of  such  accounting  items  must  again  be  borne  in  mind.     Already 
in  1930  it  was  negative  to  the  amount  of  4,110  millions.     It  was  minus 
6,040  millions  in  1931,  minus  6,550  millions  in  1932,  and  minus  3,060 
millions  for  1933.3     "Business  Savings"  as  measured  by  the  Depart- 

1  See  its  Monthly  Review  for  June  1,  1935. 

2  See  S.  Kuznets,  National  Income,  1919-1935,  193,  p.  8,  Table  I,  col.  4. 

8  See  S.  Fabricant,  Measures  of  Capital  Consumption,  National  Bureau  of  Economic 
Research  Bulletin  for  June  30,  1936,  and  Statistics  of  Income. 


THE  WORLD  CRISIS  AND  AFTER  951 

ment  of  Commerce  (National  Income,  1920-1936,  1937)  were  minus 
4,903  millions  in  1930,  minus  8,052  millions  in  1931,  minus  8,942  millions 
in  1932,  and  only  in  1935  reached  a  modest  positive  value  for  which  year 
Professor  Kuznets*  "net  savings  of  enterprises"  are  still  at  minus  3,252 
millions.1  The  fact  that  the  minima  mostly  (not  for  national  income 
measured  in  current  dollars,  however,)  occur  for  1932  is  valueless  for  us, 
because  there  can  be  little  doubt  that  government  action  in  1933  is 
responsible  for  that.2  Otherwise  those  data  are,  whatever  their  short- 
comings, full  of  interest  for  us. 

For  the  moment  the  situation  that  faced  the  incoming  administration 
certainly  was,  and  for  the  future  it  looked,  untenable.  The  unpopular 
necessity  of  refinancing  corporate  business  becomes  particularly  obvious 
if,  accepting  again  the  National  Bureau's  method  of  correcting  corporate 
accumulation  by  subtracting  the  difference  between  depreciation  on  a 
cost  basis  and  depreciation  on  the  basis  of  current  prices — which  still 
leaves  out  of  account  inadequate  provision  for  obsolescence — we  realize 
that  the  sum  total  of  those  accumulations  for  the  period  from  1919  to 
1933  turns  out  to  be  minus  7,110  millions.3  This  does  not  mean  that  the 
sum  total  of  cash  items  fell  spectacularly.  On  the  contrary,  cash  was 
precisely  the  item  that  actually  shrank  least  as  compared  with  1929 — 
from  about  7.5  to  about  6.1  billions  in  1932 — although  omission  of  current 
revaluation  formally  also  kept  up  others;  needless  to  repeat  that  this 
phenomenon  was  wholly  consequential  and  merely  reflected  but  did  not 
cause  the  spiral.  Nor  does  that  mean  that  dividends  fell  as  much  as 
earnings  which  (net  income  as  percentage  of  capitalization)  were  minus 
0.6  per  cent  in  1931  and  minus  2.8  in  1932,4  only  utilities,  foods,  beverages, 
tobacco  products,  chemical  and  allied  products,  and  (substantially) 
printing  and  publishing,  staying  on  the  positive  side  throughout.6  Not 

1  Professor  Kuznets,  op.  cit.,  Appendix  B,  discusses  these  differences,  the  numerically 
most  important  causes  of  which  are  the  governmental  dissavings  and  the  adjustments  for 
changes  in  the  valuation  of  inventories  included  in  his  estimate  but  not  in  that  of  the 
Department  of  Commerce.     In  his  series  of  net  capital  formation  in  business  at  1929 
prices  (op.  cit.>  Table  13,  p.  48,  row  II  1  6),  which  we  have  used  already,  negative  values 
begin  in  1931   with  458  millions.     1932  displays  the  minimum  of  minus  2,600  millions 
and  the  figure  for  1935  is  still  negative. 

2  That  statement  will,  however,  have  to  be  qualified  later  on. 

3  S.  Fabricant,  op.  cit.,  p.  12.     As  has  been  pointed  out  in  the  preceding  chapter,  nothing 
follows  from  that  except  the  necessity  for  the  individual  firm  to  make  accumulations  which 
may  over  time  prove  not  to  be  accumulations  at  all.     We  could,  for  example,  not  argue 
that  it  disproves  any  oversaving  theory.     On  the  contrary,  the  fact  itself  could  equally  be 
invoked  in  verification  of  the  theory  that  individual  attempts  to  accumulate  will  precisely 
result  in  the  dissipation  of  still  greater  amounts  (Ezekiel,  Keynes). 

4  Since  net  income  includes  other  elements  besides  profits  within  our  meaning  of  the 
term,  the  latter  were  negative  to  an  unknown  but  obviously  very  much  larger  percentage. 

6  National  averages  hide  many  details  which  would  shed  light  on  our  process.    Thug 


952  BUSINESS  CYCLES 

only  stockholders  in  many  corporations  but,  if  considered  as  a  class,  all 
stockholders  were  to  a  considerable  extent  allowed  to  live  on  their 
capital.  Thus  already  for  1930  negative  accumulation  ensued  from  the 
payment  of  a  total  of  net  cash  dividends  amounting  to  nearly  5.7 
billions,  while  all  the  net  income  that  remained  after  tax  payments  was 
less  than  1.3.  And  in  1931  4.2  billions  of  dividends  compare  with  a  net 
deficit  plus  taxes  of  over  3.2.  While  we  shall  think  about  the  long-run 
effect  of  this  according  to  the  theory  of  accumulation  or  saving  that  we 
make  our  own,  we  cannot  differ  about  the  remedial  or  contraspiral  effects 
such  a  behavior  must  have  had  in  the  short  run — however  much  they 
may  have  been  overcompensated  by  other  factors — particularly  since  it 
preceded  the  setting  in  of  deep  depression.  Flotations  of  new  secuities 
accord  with  those  contours.  Corporate  issues,  foreign  included,  were 
still  1,736  millions  in  1931  but  only  325  millions  in  1932,  while  municipal 
borrowing  was  active  and  the  Federal  government  borrowed  over  3 
billions  net. 

Industrial  pay  rolls  fell,  of  course,  more  strongly  than  employment 
and  arrived,  in  the  middle  of  1932,  at  about  40  per  cent  of  the  1923  to 
1925  averages.  Per  cent  fall,  less  steep  in  1931  than  it  had  been  in  1930, 
then  came  to  a  halt.  After  declining  substantially  from  the  middle  of 
June  to  the  middle  of  July,  practically  in  all  manufacturing  industries  and 
many  others  (the  main  exception  being  the  woolen-goods  industry), 
aggregate  factory  wage  payments  increased  though  subseasonally  in 
August  and  more  significantly  in  September  and  October,  after  which 
they  dipped  again  to  a  new  low  point  in  February-March  1933.  We 
interpret  as  in  the  case  of  debits.  The  German  development  was 
similar:  the  sum  total  of  wages  and  salaries  exclusive  of  pensions  but 
inclusive  of  salaries  of  public  employees  had,  according  to  the  estimate 
of  the  Institut  fiir  Konjunkturforschung,  fallen  from  its  maximum  of 
about  44.5  billion  marks  in  1929  to  about  41  billions  in  1930,  and  con- 
tinued to  fall  to  33.5  billions  in  1931  and  25.9  billions,  the  minimum,  in 

1932.  No  fall  occurred  within  the  latter  year,  and  the  figure  for  1933  is 
slightly  higher,  but  lower  for  the  first  half  as  compared  with  the  first  half 
of  1932.     In  both  countries,  therefore,  real  wage  bill  (pay  roll  by  cost-of- 
living  index)  fell  considerably.     Average  per  capita  weekly  earnings  in 
the  United  States,  as  per  monthly  data  of  the  Bureau  of  Labor  Statistics, 
declined  from  1929  to  1932  by  about  one-third  in  manufacturing,  only 
insignificantly  with  public  utilities  and  only  by  12.5  per  cent  in  retail  and 

the  textile  industry  earned  minus  6  per  cent  as  early  as  1980,  minus  6.4  in  1931,  minus  8  in 

1933,  which  places  it  near  the  head  of  the  list  of  losers.     But  if  New  England  be  excluded, 
its  place  shifts  to  about  the  average.     We  see  here  with  particular  clearness  the  connection 
of  depression  with  the  competing-down  process. 


THE  WORLD  CRISIS  AND  AFTER  953 

wholesale  trade.  *  The  fall  came  about  at  a  percentage  rate  that  increased 
to  193£  and  continued  in  1933  but  at  a  decreasing  rate,  the  total  reduc- 
tion in  the  end  amounting  to  about  36  per  cent  in  money  and  about  16 
per  cent  in  real  terms. 

Hourly  rates  fell,  to  the  middle  of  1933,  but  very  much  less — in  some 
industries,  such  as  anthracite  coal  mining,  not  at  all.  In  manufacturing 
they  declined  from  59  cents  in  1929  to  50  in  1932  and  49  in  1933.2  This 
would  yield  a  gain  in  real  terms,  and  so  would  to  a  lesser  extent  the  course 
of  money  wages  of  unskilled  labor  as  recorded  by  the  Bureau  of  Public 
Roads.  But  no  estimate  which  aims  at  a  single  figure  of  nation-wide 
significance  can  possibly  yield  a,  fall  in  real  rates.  In  Germany,  available 
quotations  of  hourly  rates  being  official  "tariff"  figures,  statistics  may 
somewhat  understate  the  fall  in  money  wages.  But  they  were,  at  the 
end  of  1932,  at  about  78  per  cent  of  the  average  of  the  maximum  year 
1930,  which  would  make  a  decline  almost  exactly  equal  to  that  in  the 
cost-of-living  index.  Now  two  things  are  obvious  from  this  behavior  of 
hourly  rates.  First,  they  cannot  have  been  a  factor  in  starting  the 
depression,  whatever  the  theory  we  may  entertain  on  the  subject:  their 
fall  cannot,  because  they  kept  up  well  at  first  and  only  reacted  to  a 
depression  already  in  full  swing;  their  previous  rise  cannot,  because  as 
we  have  seen  before,  it  was  altogether  inadequate  to  produce  that  result. 
Second,  if  there  is  anything  at  all  in  the  view  which  has  been  discussed  in 
the  preceding  chapter,  viz.,  that  the  long-run  level  of  American  money 
wage  rates,  as  distinguished  from  their  cyclical  variations,  was  "too 
high"  in  the  sense  that  it  was  partly  responsible  for  the  unemployment 
of  the  twenties,  then  it  is  clear  that  the  fall  which  occurred  during  the 
depression  was  inadequate  to  correct  that  level,  although  the  latter  might 
have  been  corrected  by  a  subsequent  rise  in  prices  occurring  without  an 
increase  in  wage  rates. 

But  it  is  more  difficult  to  say  whether  wage  rates,  by  behaving  as  they 
did,  intensified  or  alleviated  the  depression.  Since  the  dominating 
factor  in  the  short-run  situations,  especially  of  "deep"  depression,  is 
the  downward  shift  of  individual  firms'  "demand  curves"  for  labor,  and 
since  many  of  them  no  doubt  become  less  elastic  in  the  process  of  shifting 
downward,  it  is  not  only  likely  that  actual  reductions  failed,  for  the  time 
being,  to  call  forth  additional  demand  for  labor  sufficient  to  raise  the 
total  wage  bill  above  what  it  otherwise  would  have  been,  and  that  greater 
reductions  would  have  still  more  completely  failed  to  do  so,  but  there 

1  Even  the  figure  for  manufacturing  is  but  an  average  with  a  very  considerable  disper- 
sion, a  fact  which  must  be  taken  into  account  in  judging  effect.     The  figures  for  other 
industrial  groups  vary  widely  from  that  and  between  each  other — the  one  for  bituminous 
coal  is  45  per  cent. 

2  Data  of  the  National  Conference  Board. 


954  BUSINESS  CYCLES 

must  also  have  been  cases1  in  which  reductions  of  rates  simply  resulted  in 
a  decrease  of  total  output  and  employment. 

It  will  be  seen,  however,  that  this  argument  progressively  loses  force 
as  the  system  approaches  the  recovery  point  and  that  beyond  it  the  oppo- 
site conclusion  suggests  itself.  Then  our  question  admits  of  a  much 
more  definite  answer.  We  still  have  the  same  classes  of  cases  before  us. 
But  their  relative  importance  changes  when  "demand  curves"  for  labor 
tend  to  shift  upward  and  to  become  more  elastic.  Resumption  or 
expansion  of  operations  begins,  as  we  know,  in  individual  spots.  It 
certainly  did  so  begin  in  the  case  before  us,  so  that  the  effect  on  prime 
costs  of  individual  firms  is  all  that  has  to  be  taken  into  account.  Firms 
try  to  resume  or  to  expand  operations  in  a  situation  which,  while  no 
longer  discouraging,  yet  does  not  offer  those  enticements — profits  in  our 
sense  or  any  of  those  gains  which  are  induced  by  the  emergence  of  profits — 
that  later  in  the  cycle  may  make  moderate  variations  of  wage  rates  a 
matter  of  indifference.  They  are  likely  to  calculate  closely.  Even  in 
the  short  run  they  have,  particularly  if  starting  afresh  after  a  shutdown, 
some  latitude  as  to  the  combination  of  factors  that  they  are  going  to 
adopt.  The  prevailing  cheapness  of  money  will  give  them  a  slant  toward 
mechanization,  which  may  be  intensified  by  an  increase  and  counter- 
acted by  the  previous  decrease  in  wage  rates.  Hence  pay  rolls  are  likely 
to  increase  faster  in  the  absence  than  in  the  presence  of  an  increase  in 
wage  rates,  as  long  as  there  is  abnormal  unemployment.  Thus  it  seems 
permissible  to  infer,  not  only  that  such  fall  in  rates  as  occurred  facilitated 
inception  of  recovery,  but  also  that  a  stronger  fall  would,  at  least  in  the 
American  case,  have  facilitated  it  still  more.2 

D.  The  United  Kingdom,  1931-1938. — In  the  preceding  section 
English  developments  were  allowed,  first  partly  and  then  completely,  to 
drop  out  of  our  picture  because  justice  could  not  have  been  done  within  it 
to  certain  features  peculiar  to  them.  We  now  turn  to  a  discussion  of  the 
few  points  that  are  relevant  to  our  subject,  and  will  at  once  carry  our 
survey  as  near  the  present  time  as  possible.  Throughout,  it  must  be 
kept  in  mind  that  England  suffered  from  the  repercussion  of  practically 
everything  that  went  wrong  anywhere  in  the  world  and  that  she  was  a 
chief  sufferer  from  the  general  "  incapsulation "  incident  to  the  world 
crisis,  though  she  was  also  the  chief  beneficiary  of  the  fall  in  the  prices  of 
raw  materials.  She  reacted  to  all  that  by  abandoning  the  policy  of  the 

1  Various  possibilities  of  this  type  have  been  pointed  out  by  Mr.  Harrod  in  his  review  of 
Professor  Pigou's  Theory  of  Unemployment,  Economic  Journal,  March  1934,  p.  28. 

2  There  is  no  paradox  in  holding  that  a  given  event  may  intensify  depression  and,  after 
depression,  facilitate  recovery.     Nor  should  there  be  any  difficulty  in  disposing  of  the  ques- 
tion that  is  likely  to  arise  if  the  above  be  read  by  itself,  viz.,  how  "demand  curves"  are  to 
start  shifting  upward  if  "incomes"  be  kept  down. 


THE  WORLD  CRISIS  AND  AFTER  955 

Gold  Standard  Act,  by  tightening  the  economic  bonds  between  herself  and 
the  Empire  (and  some  other  countries)  and,  last  but  not  least,  by  speeding 
up  that  shift  of  her  resources  toward  production  for  the  home  market,  of 
which  the  most  important  result,  as  well  as  symptom,  was  the  Building 
Boom. 

1.  The  event  that  made  England  drop  out  of  the  line  followed  by 
our  narrative — or  would,  at  all  events,  have  caused  the  reader  to  think 
that  she  did — was  the  suspension  of  gold  payments  by  the  Bank  on  Sept. 
20,  1931,  sanctioned  the  next  day  by  the  Gold  Standard  Suspension  Act. 
The  world  depression  had  merely  made  it  more  evident  than  it  had  been 
before,  that,  under  the  social  and  economic  conditions  prevailing  at  home 
and  abroad,  the  maladjustment  caused  by  the  return  to  the  prewar  gold 
parit  would  not  disappear  of  itself  and  that,  those  conditions  being 
what  they  were,  there  was  little  if  anything  to  be  gained  by  fighting  for 
that  parity,  while  at  the  same  time  the  sacrifices  and  internal  struggles 
this  would  entail  showed  in  their  true  dimensions.  This  being  so,  it  is 
ex  post  easy  to  understand  that  abandonment  of  the  gold  standard  was 
really  a  foregone  conclusion  when  the  state  of  the  Bank's  stock  of  gold 
came  in  to  play  the  role  of  Peel's  potato  disease.1  It  had  been  well  main- 
tained, thanks  to  exceptionally  favorable  circumstances  into  which  we 
need  not  enter,2  during  most  of  1930  and  in  fact  replenished  after  Septem- 
ber 1929.  But  in  November  1930  the  efflux  set  in  which  was  to  prove 
decisive.  Government  and  Bank  acted  in  a  way  which  easily  lends  itself 
to  hostile  comment  both  from  advocates  and  from  opponents  of  aban- 
donment, but  which,  whatever  the  actual  intentions  may  have  been,  was 
"objectively"  eminently  wise.  In  order  to  take  the  plunge,  and  to 
reverse  the  policy  of  the  Gold  Standard  Act,  with  an  unbroken  front  and 

1  The  analogy  goes  pretty  far.     The  Irish  potato  plague  and  the  distress  caused  by  it 
may  have  impressed  the  Duke  of  Wellington  and  other  members  of  the  Peel  cabinet;  but  its 
effects  were  not  remedied  by  the  repeal  of  the  import  duties  on  cereals  but  by  direct  relief — 
the  "Queen's  pay" — which  would  have  been  no  less  possible  had  protection  to  agriculture 
been  retained.     Similarly,  the  loss  of  gold  did  not  absolutely  force  the  hands  of  the  govern- 
ment or  the  Bank.     But  it  provided  an  argument  that  facilitated  the  difficult  tasks  of  tran- 
sition.    The  writer  has  never  been  able  to  understand  why  some  Englishmen  should  resent 
this  suggestion  and  insist  on  England's  having  been  forced  off  gold  by  dire  and  ineluctable 
necessities.     Social  and  economic  data,  in  England  and  the  world  at  large,  being  what 
they  were,  going  off  gold  was  an  economically  rational  thing  to  do.     There  is  no  offense  in 
stating  this,  as  far  as  the  writer  can  see,  and  further  than  this  he  does  not  go.     The  author 
who  wrote  a  book  about  the  "tragedy"  of  the  pound  must  have  a  concept  of  that  literary 
genus  entirely  different  from  the  present  writer's.     The  ability  of  so  shaping  events  that 
every  action  seems  to  arise  out  of  objective  necessity  and  to  embody  the  only  possible 
course  to  take,  makes  the  unique  greatness  of  English  statesmanship. 

2  Among  them  were  gold  imports  from  countries  the  currencies  of  which  were  in  extremis 
and  which  had  to  part  with  their  gold,  such  as  Japan  and  Argentina.     Within  the  Empire, 
Australia  was  in  the  same  situation, 


956  BUSINESS  CYCLES 

to  disarm  domestic  criticism — of  which,  in  fact,  there  was  very  little 
and  none  that  counted  politically1 — it  was  first  necessary  to  do  for  the 
pound  all  that  anybody  could  ask  without  being  voted  unreasonable  by 
public  opinion.  So  the  Bank  borrowed  50  million  pounds  in  this  country 
and  France  (Aug.  7,  1931)  and  the  Treasury  announced  a  similar  trans- 
action in  the  amount  of  80  million  pounds  on  Aug.  28.2  The  Bank, 
moreover,  in  July  had  already  raised  its  rate  in  two  steps  from  2.5  to 
4.5  per  cent,  and  before  that,  in  January  1931,  had  begun  to  reduce  the 
governments  held  by  the  banking  department. 

This  raises  the  questions,  very  natural  from  the  standpoint  of  defen- 
ders of  "orthodox"  views  on  currency  and  banking,  why  both  measures 
were  not  taken  earlier — those  defenders  would  say  "in  time" — and  why, 
in  particular,  the  Bank  stayed  at  2.5  per  cent  from  May  14  to  July  23, 
i.e.,  during  the  really  critical  time.  The  rates  of  the  New  York  Federal 
Reserve  Bank  (1.5  per  cent)  and  of  the  Bank  of  France  (2  per  cent)  give 
at  best  a  partial  answer.  More  important  is  the  consideration  that  only 
a  very  modest  effect  could  have  been  expected  in  the  existing  situation. 
But  since  some  impression  could  no  doubt  have  been  made  on  gold  move- 
ments by  vigorous  handling  of  the  tools  of  central  bank  policy,  another 
conclusion  imposes  itself:  while  perfectly  ready  to  undergo  sacrifices  such 
as  are  incident  to  borrowing,  the  Bank  was  not  then  ready,  for  the  sake 
of  the  pound,  to  exert  significant  pressure  on  the  domestic  organism. 
The  timing  and  dosing  of  the  open-market  operations  mentioned,  which 
absorbed  only  funds  that  would  have  been  idle  in  any  case,3  is  conclusive 
on  this  point.  Technically,  however,  the  pound,  unlike  the  dollar,  was 
"pushed  off  gold"  and  "went  down  fighting."  And  the  sensation  all 
over  the  world — Continental  bankers  and  economists  could  not  have 
been  more  completely  stunned  by  the  news  that  Providence  had  defaulted 
on  its  bonds — was  all  the  greater. 

But  in  England  there  was  neither  panic  nor — precisely  owing  to  the 
way  in  which  the  thing  had  been  done  or,  if  the  reader  prefer,  had  come 
about — loss  of  "confidence,"  but  rather  a  sigh  of  relief.  However, 
everything  was  done  to  keep  up  discipline  and  to  prove  urbi  et  orbi  that 

1  There  were,  no  doubt,  many  individuals  who,  regardless  of  economic  interests  and 
effects,  sincerely  grieved  over  what  to  them  seemed  a  lowering  of  the  flag.    The  word 
ignominious  actually  found  its  way  into  the  pages  of  the  Economic  Journal  and  in  private 
conversation  it  may  have  been  said  that  "the  Bank  of  England  went  bust."     But  these 
feelings  did  not  amount  to  anything.     It  might  have  been  different  if  the  Labor  party  had 
been  still  in  power.     This  was,  however,  avoided  by  admirable  statesmanship  on  the  one 
side  and  no  less  admirable  discipline  and  patriotism  on  the  other:  no  party  question  arose. 

2  Both  credits  were  very  promptly  repaid — by  the  Bank  on  Oct.  81,  1981,  and  Feb.  1, 
1932;  by  the  Treasury  on  Mar.  4,  Mar.  29  and  Apr.  5,  1932. 

3  This  particular  point,  we  have  the  satisfaction  to  state,  is  recognized  by  Mr.  Hawtrey, 
op.  rit.,  p.  33. 


THE  WORLD  CRISIS  AND  AFTER  957 

this  was  not  to  be  the  beginning  of  loose  finance  and  that  there  was  no 
warrant  whatever  for  wicked  Continentals  to  talk  about  South  American 
analogies.  Bank  rate  was  put  up  to  6  per  cent  on  the  day  of  suspension 
and  actually  dwelt  on  that  vertiginous  peak  for  about  5  months — it  was 
first  reduced  to  5  per  cent  on  Feb.  18, 1932,  but  3  per  cent  was  not  reached 
until  Apr.  21;  on  June  30,  2  per  cent  came,  to  stay  to  the  time  of  writ- 
ing— dealings  in  foreign  exchange  were  temporarily  (till  Mar.  2,  1932) 
subjected  to  restriction,1  a  temporary  embargo  on  capital  issues  (relaxed 
Aug.  30,  1932)  was  declared  and,  most  important  of  all,  the  orthodox 
principles  of  public  finance  were  upheld  by  truly  heroic  efforts — an 
economy  bill  among  them,  which  was  passed  by  the  Commons  eight  days 
after  the  suspension  act — which  availed  to  end  the  financial  year  1931— 
1932  with  a  surplus  of  32.9  million  pounds.2 

From  our  standpoint  it  is  necessary  to  notice  that  this  policy,  however 
admirable  some  of  us  may  think  it  was,  exerted  some  pressure  on  the 
economic  process  and,  for  the  time  being,  tended  to  neutralize  such 
stimulating  effects  as  the  depreciation  of  the  pound  might  have  had. 
This  statement  implies  the  admission  that  depressing  effects  may  in 
particular  have  emanated  from  the  6  per  cent  bank  rate.  Personally, 
the  writer  doubts  whether  a  single  additional  pair  of  boots  would,  during 
those  5  months,  have  been  produced  had,  other  things  being  equal,  bank 
rate  been  lower.  Still,  since  the  rate  might  easily,  i.e.,  without  producing 
per  se  any  very  catastrophic  consequences,  have  been  3  per  cent  instead 
of  6,  and  since  it  is  no  part  of  our  theory  of  central  banking  to  hold  that 
a  difference  as  considerable  as  this  is  a  matter  of  no  importance  to 
producers'  business,  we  will  for  argument's  sake  concede  that  its  pressure 
did  extend  beyond  the  open  market,  provided  that  the  reader,  in  exchange, 
be  good  enough,  first,  to  observe  on  any  of  the  familiar  graphs  that  busi- 

1  Those  restrictions  were,  in  April  1982,  replaced  by  the  establishment  of  the  Exchange 
Equalisation  Account — misleadingly  called  a  Fund — which  serves  the  same  purpose  more 
effectively  and  other  purposes  besides.     Even  from  the  standpoint  of  our  subject,  this 
original  gadget  would  deserve  more  attention  than  we  can  afford  to  bestow.     It  must  suffice 
to  note  that  its  operation  may  powerfully  intensify  or  offset  the  open-market  operations 
of  the  Bank  and  be  quite  as  effective  in  regulating  the  money  market  as  these,  though  much 
more  difficult  to  observe.     If  the  Fund  simply  bought  and  sold  gold  and  foreign  currencies, 
the  effect  on  member  banks*  cash  and  reserve  with  the  Bank  would  be  much  the  same  as  the 
effect  of  gold  movements;  the  only  difference  would,  in  this  case,  consist  in  that  the  Fund's 
operations  are  planned.     This  effect,  however,  may  be  counteracted  by  concomitant  sales 
and  purchases  of  treasury  bills;  but  again,  it  may  not.     If  the  Fund  sells,  say,  dollars  and 
deposits  the  proceeds  with  the  Bank  without  relieving  member  banks  of  a  corresponding 
amount  of  treasury  bills,  a  restrictive  influence  would  be  exerted.     Hence,  we  have  here  a 
new  organ  of  central  banking,  which  would  be  very  important  for  us  if  we  could  go  into 
details. 

2  The  official  result  was  a  surplus  of  only  0.4  millions.     But  "expenditure"  contains  an 
item  of  32.5  millions  for  sinking-fund  purposes. 


958  BUSINESS  CYCLES 

ness  was  not  more  depressed  during  those  5  months  than  we  should  have 
expected  it  to  be  in  any  case — the  quarterly  index  of  production  of  the 
London  and  Cambridge  Service  first  rose  and  then  stayed  up  during 
those  critical  months — and,  second,  to  admit  that  if  there  was  any 
aftereffect  later  in  193£  it  cannot  have  been  very  terrible,  since  by 
August  the  building  boom  had  set  in.  However,  together  with  all 
other  measures  taken — in  particular,  the  balancing  of  the  budget — this 
policy  no  doubt  substantially  contributed  to  all  those  features  of  the 
English  process  that  make  so  striking  a  contrast  with  the  course  of 
events  in  the  United  States:  a  speculative  flare-up  and  the  consequent 
relapse  were  prevented,  expansion  was  patiently  awaited,  secure  foun- 
dations were  laid  for  the  sustained  and  sober  advance  that  was  to  last 
with  practically  no  interruption  until  the  spring  of  1938. 

2.  To  complete  this  part  of  our  sketch,  as  soon  as  it  felt  safe  ground 
under  its  feet,  the  Bank  made  a  move  toward  monetary  expansion,  and 
along  with  reducing  its  rate  from  6  to  2  per  cent,  bought  government 
securities.1  The  immediate  object  was  preparation  for  the  great  refund- 
ing and  conversion  operations  of  that  year.2  But  the  ultimate  effect  was 
to  end  the  period  of  easy  money  at  high  rates — no  reason  to  object  to 
this  as  a  paradox — and  to  usher  in  the  period  of  easy  money  at  low  rates. 
Day-to-day  money,  for  instance,  which  in  the  average  of  1929  had  been 
at  4.57,  fell  to  0.68  for  the  third  quarter  of  1932,  then  was  0.66  per  cent 
in  the  average  of  1933,  0.81  per  cent  in  the  average  of  1934,  0.73  per  cent  in 
the  average  of  1935,  and  from  April  of  that  year  stayed  put  at  0.75  per 
cent  to  the  time  of  writing — the  manometer  of  interest  was  paralyzed 
almost  as  effectively  as  in  the  United  States. 

Those  open-market  operations,  however,  kept  within  sober  bounds. 
The  annual  average  of  government  securities  held  in  both  departments  of 
the  Bank,  which  in  1930  and  1931  had  been  about  295  million  pounds, 
was  only  312  for  1932  and  335  in  1933.  But  this  and  the  influx  of  gold 
helped  to  swell  the  cash  reserves  of  member  banks  (London  clearing 

1  In  doing  so,  it  not  only  offset,  but  during  1932  occasionally  more  than  offset,  losses  of 
gold.     See  on  this  and  cognate  subjects,  S.  E.  Harris,  British  and  American  Exchange 
Policies,  I,  The  British  Experience,  Quarterly  Journal  of  Economics  for  May  1934. 

2  It  must  again  be  borne  in  mind  that  the  money  market  and  the  banking  policy  were, 
all  along,  dominated  by  the  requirements  and  actions  of  the  Treasury  and  by  the  mechanics 
of  public  finance.     Release  of  government  loan  dividends,  provision  for  application  money 
for   government   issues,   income  tax  payments,   treasury  preparations   for   heavy  pay- 
ments, and  operations  to  facilitate  all  this  or  to  offset  the  effects  of  it  all  were  so  important 
that  a  history  of  the  English  money  market  during  those  years  could  almost  be  written  in 
terms  of  them.     This  is  another  important  subject  that  must  be  dismissed  in  a  note.     That 
open-market  operation  and  the  conversion  of  2,086  million  pounds  of  5  per  cent  war  loans 
to  3.5  per  cent  (announced  June  30,  1932)  were,  of  course,  necessary  in  order  to  remove  an 
obstacle  on  the  road  to  cheap  money. 


THE  WORLD  CRISIS  AND  AFTER  959 

banks,  average  in  1931,  182  millions;  1933,  212)  and  their  investments 
(London  clearing  banks,  annual  average  of  1931,  301  millions;  of  1932, 
348;  of  1933,  537;  of  1934,  560;  of  1935,  615;  no  increase  in  1936;1  small 
increase  in  1937;  decrease  in  the  first  quarter  of  1938). 

This  increase  in  investments  is,  of  course,  reflected  in  deposits, 
especially  in  their  great  stride  in  1932—1933,  which,  however,  did  not 
fail  to  be  accompanied  by  a  symptom  of  the  truth  that  banks*  invest- 
ments tend  to  produce  idle  deposits;  for  the  ratio  of  current  accounts  to 
total  deposits,  which  was  about  54  in  1929  and  then  naturally  fell — we 
remember  that  there  was  no  fall  in  total  deposits  in  1930;  the  fall  in  1931 
was  insignificant,  all  the  loss  of  gold  notwithstanding,  which  shows  again 
that  also  in  this  case  monetary  stringency  cannot  have  been  a  major 
element  in  the  depressive  processes — reached  its  minimum  (50)  for  1932 
and  stayed  near  its  minimum  in  1933  (a  little  over  5 1).2  Total  deposits 
in  the  nine  clearing  banks  were  above  the  1929  figure  already  in  1932 — 
the  further  fall  in  current  accounts  that  occurred  in  that  year  being  more 
than  compensated  by  an  increase  in  deposit  (time)  accounts — and,  1,914 
millions  in  the  average  of  1933.  The  slight  setback  in  business  that 
occurred  in  1934  is  reflected,  in  spite  of  continued  increase  (over  the 
year)  of  investments,  in  a  small  reduction,  but  the  average  for  1936  was 
1,961 — 200  millions  above  1929 — the  average  for  1936  over  2,100  and 
that  for  1937  still  higher.  No  shrinkage  but,  on  the  contrary,  further 
increase  occurred  in  1938  (to  September). 

To  the  last  quarter  of  1935,  advances  contributed  next  to  nothing  to 
this  expansion  of  deposits.  We  have  here  the  same  phenomenon  that 
will  have  to  be  mentioned  in  the  American  and  the  German  case — a 
similar  structural  change  in  bank  assets.  Advances  shrank  sharply  in 
1932  and  kept  creeping  along  at  about  the  same  level  for  over  two  years 
afterward.  But  they  increased  significantly — as  they  should  within  the 
rhythm  of  our  schema — from  the  fourth  quarter  of  1935  to  the  first 
quarter  of  1938,  and  for  part  of  this  period  dominated  variations  in 
deposits.  The  same  is  true  of  that  part  of  discounts  which  consists  of 
commercial  bills,  if  we  may  judge  from  the  returns  of  the  three  banks  that 
publish  these  figures  separately.  Otherwise,  the  bulk  of  the  portfolios 
consisting  of  treasury  bills,  the  discount  item  is  not  of  much  help  to  us. 

3.  Those  effects  of  monetary  management  as  practiced  in  England 
which  do  not  work  via  international  relations  may,  therefore,  be  summed 
up  by  stating — what  will  presently  be  verified  by  time-series  evidence — 
that  it  brought  out  rather  than  obliterated  the  features  of  the  economic 

1  This  is  the  "trace"  left  of  the  "normal"  movement  of  member  bank  investments  in 
the  cycle,  since  1936  may  be  considered  as  part  of  a  Juglar  prosperity. 

2  In  1935  it  climbed  up  to  what  was  nearly  the  1929  figure — another  trace  of  the  cyclical 
regularity. 


960  BUSINESS  CYCLES 

process  which  we  should  have  expected  to  observe  in  the  "normal"  course 
of  things.  Availing  itself  of  the  freedom  gained  by  the  disruption  of 
the  galling  gold  ligamen,  it  no  doubt  protected  that  process  from  what 
would  otherwise  have  happened  to  it  and  avoided  the  painful  operations 
which,  without  going  off  gold,  would  have  been  necessary.  But  it 
neither  aimed  at  nor  effected  anything  that  could  in  any  useful  sense  be 
called  an  "inflationary"  impulse,  which  in  fact  is  not  necessarily  inherent 
in  either  devaluation  or  depreciation.  It  (and  concomitant  policies) 
not  only  did  not  aim  at  but,  if  anything,  discouraged  income  generating 
public  expenditure  and  increase  in  rates  of  earnings  or  commodity  prices. 
Thus  it  was  more  remarkable  and  successful  by  what  it  refrained  from 
doing  than  by  what  it  did.  The  effects  which  worked  through  inter- 
national relations  must  be  viewed  in  their  place  within  a  comprehensive 
policy  of  reorientation. 

a.  The  relief  immediately  given  to  foreign  trade  and  conserved  by  the 
stability  of  money  wage  rates  and  domestic  prices  is  likely  to  be  under- 
estimated1 because  it  mainly  consisted  in  warding  off  much  of  that 
additional  shrinkage  which  would  otherwise  have  ensued,  at  least  during 
1932.  There  was,  hence,  no  good  reason  for  the  disappointment  expressed 
at  the  apparent  absence  of  any  considerable  immediate  effects.  Also, 
the  relief  was  in  part  counteracted  by  measures  taken  in  other  countries, 
both  in  retaliation  to  and  independently  of  the  English  move.2  As  far  as 
these  reactions  did  not  consist  in  tightening  quota  and  so  on  but  in  sym- 
pathetic depreciation  of  currencies,  it  must,  however,  be  emphasized 
that  they  did  not  entirely  defeat  the  English  purpose  even  in  the  sphere 
of  international  trade.  For  wherever  the  gold  ligamen  galled,  its  removal 
would  have  helped  English  trade  by  improving  the  general  situation  in 
the  respective  countries,3  and  its  removal  coupled  with  that  spending 
policy  from  which  England  substantially  refrained,  would  even  have  given 
England  a  relative  advantage  in  addition  to  the  absolute  one.  Because 

1  Methods  are  available  for  arriving  at  a  rough  estimate  of  that  effect.     For  our  purpose 
it  is,  however,  not  necessary  to  enter  into  what  would  be  a  very  laborious  and  expensive 
investigation. 

2  There  really  was  a  third  category:  without  wishing  to  retaliate,  foreign  countries  found 
it  easier  to  abandon  the  gold  standard  and  to  revise  their  promises  to  their  creditors  because 
the  procedure  had  been  made  respectable  by  the  English  example  and  no  longer  carried  the 
connotation  it  used  to.     The  implications  of  this  fact  are  most  important  for  the  general 
theory  of  morals.     Japan,  for  instance,  who  went  off  on  Dec.  14,  1931,  had  no  doubt  very 
good  reasons  for  doing  so.     But,  attaching  much  importance  to  her  impeccable  financial 
record,  she  would  presumably  have  hesitated  to  take  the  step  (though  it  might  have  become 
inevitable  later  on)  without  the  precedent  created  three  months  earlier. 

8  It  is,  hence,  not  correct  to  say,  as  was  said  at  the  time,  that  England  tried,  by  abandon- 
ing the  gold  standard,  "to  push  the  burden  of  depression  on  to  other  countries."  Even 
the  element  of  truth,  however,  which  is  contained  in  this  phrase,  should  in  fairness  read 
that  she  tried  to  push  back  the  burden  that  had  first  been  pushed  on  to  her. 


THE  WORLD  CRISIS  AND  AFTER  961 

of  this  and  other  reasons,  it  was  perfectly  logical  to  refuse  to  consider 
stabilization  of  the  pound  at  any  particular  value,  and  the  permanent 
refusal  to  do  so  would  be  an  understandable  policy. 

Two  other  factors  must,  moreover,  be  taken  into  account.  First, 
as  has  been  stated  before,  the  great  fall  in  the  value  of  exports  of  manu- 
factures had  occurred  from  the  last  quarter  of  1929  to  the  beginning  of 
1931.  During  the  latter  year,  the  rate  of  shrinkage  rapidly  decreased 
throughout  the  three  quarters  preceding  suspension;  and  in  spite  of 
depreciation,  the  monthly  average  for  1932  was  still  lower,  though  no 
doubt  higher  than  it  would  have  been  without  it.  The  average  for  1933 
(23.4  millions)  was  very  little  higher  and  even  the  average  for  1936  (28.4) 
was  but  55  per  cent  of  the  average"  for  1924,  or  59  per  cent  of  the  average 
for  1929.  Only  the  first  half  of  1937  showed  a  vigorous  upswing  (season- 
ally corrected  figure  for  June:  37  millions)  which  then  gave  way  to 
relapse  (June  1938:  30.7  millions,  which  was  better,  however,  than  the 
values  for  April  and  May) .  But  this  contour  simply  describes  the  course 
of  (English  and  foreign)  cyclical  phases  and  is  what  on  a  somewhat  lower 
level  we  should  have  expected  to  observe  in  any  case.  The  conclusion 
which  follows  for  the  degree  of  influence  exerted  by  suspension  is  obvious. 
Furthermore,  part  of  the  success  attained  must  be  attributed  to  adap- 
tations to  new  conditions  which  would  have  asserted  themselves,  depre- 
ciation or  no  depreciation:  while  the  immediate  effects  of  depreciation, 
in  1931  and  1932,  were  particularly  noticeable  in  the  exports  of  textiles 
to  the  Far  East — British  exporters  making  promptly  concessions  in  the 
foreign  prices  of  their  commodities1 — the  increase  which  occurred  in 
1934,  for  example,  was  only  to  15  per  cent  due  to  textiles,  which  tended 
to  drop  into  the  background  while  other  articles  were  coming  to  the  fore. 
This  adaptation  or  shift  was  no  doubt  facilitated  or  conditioned  by  the 
depreciation  of  the  pound,  but  cannot  be  called  simply  its  automatic 
consequence. 

Second,  suspension  was  accompanied  by  England's  definitive  con- 
version to  neomercantilism.  Her  monetary  policy  is  in  fact,  but  the 
complement  of  it. 

6.  The  Abnormal  Importations  Act  (Nov.  30,  1931;  three  lists  of 
emergency  duties  followed  within  the  year)  which  marks  the  transition, 
presented  itself  as,  and  in  a  sense  undoubtedly  was,  an  emergency 
measure.  It  added  to  the  effectiveness  of  depreciation  in  curtailing 
imports  and  in  relieving,  on  balance,  the  immediate  situation.  The  two 
measures  combined  account  for  the  bulges  in  all  wholesale  price  indices, 
as  well  as  in  the  indices  of  prices  of  foods,  nonfoods,  and  materials,  which 
occur  for  the  last  quarter  of  1931  and  were  eliminated  or  more  than 
eliminated  by  the  middle  of  1932.  But  excepting  some  effects  on  specu- 
1  See  Professor  Harris  op.  cit.,  p.  487. 


962  BUSINESS  CYCLES 

lative  anticipations,  depreciation  did  not  immediately  affect  domestic 
prices — or,  for  that  matter,  business  activity — in  any  way  other  than  by 
the  prices  and  quantities  of  English  foreign  trade. 

The  frank  adoption  of  the  principle  of  (moderate)  protection  by  the 
tariff  act  of  1932  acquires  its  historic  importance  and  the  glamour  of  vast 
possibilities  in  connection  with  the  11  bilateral  agreements  between  the 
different  parts  of  the  Empire  (passed  by  the  House  of  Commons  on 
Nov.  3,  1932)  which  gave  effect  to  the  understandings  arrived  at  in 
Ottawa  (July  21-Aug.  20,  1932)  and  crowned  the  efforts  of  more  than 
thirty  years.  It  does  not  follow,  however,  that  the  importance  of  these 
achievements  is  equally  great  for  our  subject  and  for  the  period  under 
survey.  In  Chap.  VI  we  have  seen  that  the  immediate  effects  on  the 
cyclical  process  of  the  German  Zollverein  were  by  no  means  commen- 
surate with  what  one  might  expect  when  judging  from  the  long-run 
developments  for  which  it  in  part  provided  the  frame.  The  same  is, 
mutatis  mutandis,  true  of  the  case  in  hand  and  from  our  standpoint  we 
can  dispose  of  it  by  three  remarks. 

First,  although  the  new  policy  affected  total  imports  of  manufactures 
considerably  and  still  more  many  individual  items,  such  as  motorcars, 
neither  depreciation  nor  protection  prevented  values  of  total  net  imports 
from  recovering,  in  1933  and  1934,  fully  as  much  as  values  of  exports  of 
United  Kingdom  produce  or  manufactures.  They  increased  more  than 
values  of  exports  in  1935,  1936,  and  especially  in  the  first  half  of  1937. 
This  means  that  the  new  policy  did  not  prevent  them  from  behaving 
according  to  the  cyclical  schema. 

Second,  it  is  worth  while  to  notice  the  contribution  of  depreciation 
to  the  Ottawa  success.  It  is  hardly  conceivable  that  things  would  have 
been,  comparatively  speaking,  so  smooth  if  England  had  had  to  appear 
in  the  role  of  the  adamant  creditor  instead  of  appearing,  as  she  was  able 
to  do,  in  the  role  of  the  understanding  friend  who  had  just  waived  a 
substantial  part  of  his  rights.  Moreover,  depreciation  and  monetary 
management  reestablished  England's  lending  power,  which  the  gold 
maladjustment  and  the  crisis  had  all  but  destroyed,  and  thus  made  her 
again  what  she  had  been  before,  the  giver  of  all  financial  delights. 

This  element  we  must  bear  in  mind  when  judging  the  balance  of 
advantages  and  disadvantages  that  accrued  to  England  from  the  depreci- 
ation of  her  currency  and  from  her  methods  of  monetary  management. 
Both  had  much  to  do  with  the  relatively  good  showing,  during  the  world 
crisis,  of  her  income  from  foreign  investments,  which  in  turn  mitigated  the 
impact  of  the  depression  on  her  domestic  process.  For  instance,  the 
"income  from  British  investments  in  Oversea  Securities  dealt  in  or  known 
to  the  London  market"1  fell  indeed  from  its  1929  figure  of  212.4  millions 

1  See  the  articles  on  the  subject  by  Sir  R.  M.  Kindersley  in  the  Economic  Journal.  The 
above  figures  are  taken  from  the  one  in  the  number  for  December  1037,  p.  654,  Table  IX. 


THE  WORLD  CRISIS  AND  AFTER  963 

to  138.3  millions  in  1933  (minimum)  and  was  only  164.4  even  in  1936, 
but  that  part  of  it  which  came  from  loans  to  Dominions,  Colonial  and 
Foreign  Governments  and  Municipalities  kept  up  well  during  the  crisis 
years,  though  it  continued  to  fall  later  for  other  reasons,  and  even  the 
total  is  much  more  comforting  to  look  at  than  it  would  otherwise  have 
been. 

Third,  the  Ottawa  policy  did  not  stop  at  the  frontiers  of  the  Empire. 
The  "sterling  bloc"  extended  beyond  and  included  several  valuable 
conquests.  Tariff  reductions  and  quota  facilities  for  English  commodities 
were,  for  example,  secured  by  treaties  with  the  countries  that  had  most 
promptly  linked  their  currency  to  the  pound — Denmark,  Norway  and 
Sweden — but  much  more  than  that  resulted  in  the  case  of  the  convention 
with  the  Argentine  Republic  (signed  May  1,  1933,  supplement  Sept.  26). 
In  this  case  the  remission  of  debt  implied  in  the  depreciation  of  the 
pound,  coupled  with  an  additional  loan,  was  made  the  cornerstone  for 
arrangements  about  the  unfreezing  of  credits,  commodity  trade,  and 
control  of  foreign  exchange  dealings  and  so  on,  in  a  way  highly  advan- 
tageous to  Great  Britain.1  It  may  take  some  enthusiasm  to  style  that 
arrangement  as  the  granting  of  Dominion  status  to  Argentina.  But  it  is 
an  outstanding  example  of  the  able  and  conscientious  handling  of  indi- 
vidual sources  of  actual  loss  and  potential  gain  which  no  doubt  greatly 
relieved  the  domestic  cyclical  process. 

4.  As  stated  above,  English  industrial  processes  displayed,  besides 
the  general  characteristics  appropriate  to  the  Kondratieff  phase  and  the 
special  features  characteristic  of  the  downgrade  of  this  Kondratieff, 
also  the  effects  of  the  reorientation  of  the  English  economy  in  response 
to  the  change  that  had  occurred  in  England's  international  position. 
The  first  class  of  phenomena  does  not  call  for  additional  comment: 
electric  power  (more  and  more  completely  developing  into  state  enter- 
prise), electric  manufacture  (wire,  cables,  installation,  lamps,  apparatus, 
and  machinery),  motorcars,  nonferrous  metals,  chemicals  including 
rayon,  aeroplanes,  and  so  on  continued  to  progress  without  much  inter- 
ruption even  in  the  worst  year,  the  British-owned  rubber  plantations 
being,  within  this  class",  the  interest  that  suffered  most  during  the  crisis. 
For  some  of  the  chief  sufferers  from  reorientation,  such  as  the  coal  and 
iron  industries,  the  situation  was  rendered  still  worse  because  the  normal 
process  of  innovation2  would  have  made  them  centers  of  depressive 
symptoms  without  that.  This  was  not  so,  however,  with  textiles  and 
shipbuilding,  although  the  latter  experienced  a  spectacular  recovery  in 
1934  and  again  in  1936,  lasting  through  1937. 

1  Cf.  R.  B.  Stewart,  Anglo- Argentine  Trade  Agreements,  Canadian  Journal  of  Economics 
and  Political  Science  for  February  1936. 

2  The  chief  instance  of  a  major  industry  that  suffered  only,  but  suffered  severely,  from 
the  effects  of  economic  "progress,"  is  supplied  by  the  railways. 


964  BUSINESS  CYCLES 

In  the  second  class  of  phenomena,  the  most  important  item,  in  fact 
the  item  that  may  with  but  little  qualification  be  said  to  have  "carried" 
recovery  and  prosperity  was  the  building  boom.1  It  was  a  boom  in  a 
consumers'  goods  industry,  dwelling  houses  supplying  from  1932  to  1934 
over  70  per  cent  of  the  value  of  building  plans  approved,  the  share  of 
factories  not  reaching  8.6  per  cent  until  1936.  And  though  mainly,  it 
was  not  only  conditioned  by  that  general  shift  toward  production  for 
home  demand  and,  as  an  aid  to  or  symptom  of  this  tendency,  by  the 
impediments  placed  in  the  way  of  foreign  investment  or  by  the  unpromis- 
ing outlook  for  investment,  either  as  to  safety  or  as  to  rate  of  return,  in 
most  foreign  countries.  Fiscal  policy,  the  lead  given  in  the  twenties 
by  the  policy  of  subsidies,  the  effect  this  subsidized  housing  had  exerted 
on  the  development  of  an  efficient  large-scale  building  industry,  the  stable 
rates  of  earnings  which  in  a  time  of  falling  cost  of  living  left  a  large  middle 
stratum  that  suffered  comparatively  little  from  unemployment  with  a 
margin  to  spend,  the  fall  of  building  costs  during  "deep"  depression,  the 
subsequent  cheap  money  policy,  industrial  migration — all  this  power- 
fully propelled  either  the  demand  for  housing  or  the  means  of  satisfying 
it  and  thus  helped  to  condition  the  boom  in  unassisted  building.  Some 
of  these  favorable  conditions  were  only  temporary.  Cost  of  living  began 
to  rise  (after  relapse  from  the  rise  in  1931-1932)  in  the  spring  of  1935  and 
then  rose  considerably  to  the  summer  of  1937.  The  index  of  building 
costs  (1924  =  100)  which  had  been  at  124.2  for  1930  and  at  107.5  for 
1931  was  again  at  124.5  in  1932  and  rose  to  165.4  for  1933.  The  mortgage 
rate  was  slow  in  falling  from  its  maximum  in  1931.  The  first  important 
reduction  occurred  toward  the  end  of  1932,  i.e.,  after  the  boom  had  started, 
and  the  effects  of  the  further  fall  in  1933  together  with  the  easing  of  other 
conditions  were  more  than  offset  by  the  increase  in  building  costs.  The 
cheap  money  policy  thus  helped  the  building  activity  by  shifting  toward 
it  funds  that  would  have  had  to  be  content,  from  the  middle  of  1932, 
with  less  than  4  per  cent  elsewhere,  and  by  thus  increasing  the  power  of 
the  specialized  machine  for  the  financing  of  home  building  rather  than 
by  offering  direct  stimulation. 

The  activity  in  unassisted  housebuilding  as  measured  by  the  number 
of  houses  completed  without  state  assistance  (Ministry  of  Health;  the 
figures  are  not  quite  complete;  years  are  from  Oct.  1  to  Sept.  30)  got 
into  its  stride  in  1930,  the  year  after  discontinuation  of  building  under 
the  Chamberlain  Act,  110,375  houses  being  completed  as  compared  with 
71,083  in  1929.  The  figure  for  1931— the  year  of  unprecedented  and 

1  See  ante,  Chap.  XIV,  Sec.  D,  2.  On  the  facts  and  the  theory  of  the  boom,  cf.  a 
forthcoming  book  by  Mr.  W.  Stolper;  for  the  rise  and  policy  of  the  Building  Societies  in 
particular  the  article  by  Sir  H.  Bellman  in  Economic  Journal  for  March  1933.  Total 
resources  of  building  societies  eventually  rose  to  over  700  million  pounds. 


THE  WORLD  CRISIS  AND  AFTER  965 

unbearable  catastrophes1 — is  132,909  and  that  for  1932,  the  worst  year 
in  our  period,  132,886.  Then  follows  an  ascent  at  increasing  rates  to 
1935  (283,453).  But  although  declining,  home  building  has  kept  going 
strong  to  the  time  of  writing.  Even  in  the  first  quarter  of  1938  the  figure 
of  value  of  housebuilding  plans  approved  was  exactly  the  same  as  that 
for  the  first  quarter  of  1937,  the  decline  in  total  value  of  building  plans 
approved  being  exclusively  chargeable  to  industrial  construction,  and  the 
figure  for  the  second  quarter  of  1938  was  but  7  per  cent  below  that  for  the 
second  quarter  of  1937.  This  boom  is  perfectly  regular  in  the  sense  of 
our  schema.  Its  natural  end  will  make  a  not  less  regular  contribution 
to  what  within  that  schema  should  be  the  depression  phase  of  the  current 
Juglar. 

Here  we  may  notice  the  role  in  the  cyclical  process  of  actual  and 
prospective  expenditure  on  armaments,  which  began  to  assert  itself  at 
just  about  the  time  when  the  building  boom  had  passed  its  peak:  though 
such  expenditure  is  the  almost  unavoidable  concomitant  of  neomercan- 
tilist  policy,  it  was  the  Abyssinian  incident  that  caused  it  to  step  out  of 
the  frame  of  the  current  budget  and  to  become  a  new  factor  of  the 
economic  situation.  It  means,  on  the  one  hand,  the  building  up  of  a  new 
industrial  structure  for  an  armament  industry  of,  at  all  events,  new 
proportions,  which  is  the  primary  desideratum.2  Whether  left  to 
private  or  effected  by  public  enterprise,  we  have  here  a  "  conditioning " 
task  which  may  be  perfectly  adequate  for  carrying  a  Juglar  prosperity, 
even  if  we  discount  fantastic  figures  inspired  by  passing  panics.  On  the 
other  hand,  this  means  a  break  in  the  government's  financial  policy,  of 
which  the  abstention  from  anything  like  American  spending — from 
income  generation  by  public  spending — was  one  of  the  major  elements 
until  them.  What  has,  for  reasons  which  the  writer  believes  were  justified 
by  results,  been  denied  to  public  works  is  now  perforce  granted  to 
armaments.  Effects  may  some  day  dominate  the  picture  to  the  complete 
extinction  of  cyclical  contours,  but  so  far  this  is  not  the  case.  For  the 

1  It  is  true  that  a  boom  in  some  small  industry  may  start  in  deep  depression,  owing  to 
circumstances  peculiar  to  that  industry,  without  affecting  the  general  picture,  which  may 
continue  to  be  one  of  deep  depression.     This  is  not  possible  in  the  case  of  a  boom  in  dwelling- 
house  building,  which  presupposes  conditions  of  demand,  and  entails  effects  on  general 
business,  incompatible  with  deep  depression.     To  speak  of  deep  depression  or  of  unrelieved 
gloom  amidst  the  beginnings  of  such  a  boom  is  an  abuse  of  terms. 

2  The  English  armament  industry  was  always  considerable,  from  the  seventeenth 
century  on;  but  its  size  was  a  function  of  foreign  rather  than  domestic  demand.     The 
development  during  the  World  War  was  to  a  considerable  extent  undone  during  the  subse- 
quent 15  years.     Exports  of  arms  (including  torpedoes  and  the  like),  ammunition,  and 
warships  amounted  in  1984  to  only  about  8.36  million  pounds,  and  this  figure  characterizes 
the  size  of  the  armament  industry  fairly  well.     There  was,  of  course,  a  quickly  expanding 
aviation  industry  in  addition. 


966  BUSINESS  CYCLES 

fiscal  year  1937-1938  the  socalled  special  rearmament  expenditure 
financed  by  loans — which  is  additional  to  the  Defense  item  of  the 
ordinary  budget — was  only  65  million  pounds,  and  it  would  take  much 
more  than  that  to  induce  what  is  euphemistically  referred  to  as  "expan- 
sion," especially  as  long  as  vigorous  taxation  and  economy — 21  millions 
were  in  the  same  year  economized  on  Civil  Expenditures — are  adhered  to. 
Some  steadying  effects  on  employment  in  recessions  and  depressions  and 
some  pressure  on  the  standard  of  life  are  all  that  can  be  immediately 
expected.1 

Budgetary  deficits,  if  any,  continued  to  be  insignificant.  So  was  the 
increase  in  the  National  Debt,  if  we  exclude  that  part  of  it  which  was 
incurred  in  order  to  finance  the  gold  purchases  of  the  Exchange  Equalisa- 
tion Account.  But  it  would  be  forcing  open  doors  if  we  tried  to  prove 
that  no  government  "inflation"  lent  its  aid  to  the  steady  progress  of 
those  years. 

5.  Evidence  from  time  series  confirms  the  impression  we  have  gathered 
so  far:  neither  new  policies  nor  the  numerous  disturbances  to  which 
England's  economy  was  exposed,  effaced  the  cyclical  contour  lines  which, 
taking  account  of  specifically  English  conditions,  we  should  expect  from 
our  schema. 

Whether  we  take  the  Board  of  Trade's  index  of  production  or  that  of 
the  London  and  Cambridge  Economic  Service,  we  cannot  fail  to  be  struck 
by  the  gentleness  of  the  descent  into,  and  the  steadiness  of  the  recovery 
from,  the  depression,  which  characteristically  distinguishes  the  English 
curve  from  any  other,  the  Nordic  countries  which  followed  the  same 
monetary  policy  not  excluded.2  This  steadiness  is  as  much  in  evidence 
before  as  it  is  after  the  abandonment  of  the  gold  standard.3  Between 
the  middle  of  1931  and  the  middle  of  1932  there  is  a  little  hump  which  is 
not  absent  but  much  less  marked  in  some  countries  that  remained  on  the 
gold  standard,  so  that  it  may,  partly  at  least,  be  attributed  to  deprecia- 
tion. The  minimum,  which  occurs  in  1932,  is  lower,  however — the  Lon- 
don and  Cambridge  Service  quarterly  index  has  77.8  (average  of  1924  = 
100)  for  the  third  quarter — and  the  recovery  of  the  fourth  quarter  then 
strongly  continued  in  1933,  without  any  appreciable  rise  in  prices,  leading 

1  The  above  paragraph  was  written  in  the  late  summer  of  1988. 

2  See  e.g.,  Chart  I  on  p.  256  of  D.  Westcott's  article,  previously  quoted,  in  the  Review  of 
Economic  Statistics  for  December  1934. 

8  That  gentleness  of  sweep — in  some  cases  approaching,  until  1983,  a  horizontal — is 
also  observable  in  the  majority  of  components,  especially  in  Foods  and  Drinks,  Leather  and 
Boots,  and  from  the  middle  of  1980  on,  in  chemicals,  textiles,  non-ferrous  metals.  Engi- 
neering including  Shipbuilding  and  Mining  displays  a  stronger  downward  slope  to  middle 
of  1932,  and  Steel  is  the  most  cyclical  series  of  all.  The  steady  upswing  in  the  latter  from 
middle  of  1932  clearly  falls  into  two  segments,  one  covering  1933,  1984  and  hah8  of  1935,  the 
rest  covering  the  second  half  of  1935,  1936  and  three  quarters  of  1937. 


THE  WORLD  CRISIS  AND  AFTER  967 

up  to  102.5  for  the  first  quarter  of  1934.  The  general  improvement  of 
things  which  is  observable  in  that  quarter  was  associated  with  a  tem- 
porary reduction  of  the  income  tax.  In  any  case,  however,  and  whatever 
the  shortcomings  of  indices  may  be,  we  have  before  us  a  broad  bottom 
extending  over  1931  and  1932  which  in  time  and  shape  exactly  corresponds 
to  the  idea  we  have  about  how  the  depression  of  a  Juglar  located  as  this 
was  should  look,  followed  by  a  perfectly  normal  recovery  phase  which  it 
would  be  highly  unreasonable  to  attribute  to  monetary  policy  as  "the" 
cause  though,  as  we  expressed  it  above,  this  monetary  policy  undoubtedly 
"protected"  it  from  the  consequences  that  adherence  to  the  gold  standard 
would  have  had.  It  is  also  perfectly  en  regie  that  recovery  tapered  off  in 
the  course  of  1934,  and  so  is  the  vigorous  upswing  in  the  last  quarter  of 
1935,  which,  had  we  a  desire  to  press  the  point,  could  be  identified  as  the 
rise  of  the  fifth  Juglar. 

The  statement  that  recovery  was  perfectly  normal  would,  however, 
require  qualification  if  we  accept  the  testimony  of  the  London  and  Cam- 
bridge Service's  quarterly  index;  for  its  figures  for  1934,  though  higher 
than  those  of  1924,  are  lower  than  those  of  any  of  the  years  from  1927  to 
1929  and  thus  fall  below  expectation.  But  we  know  that  this  is  due  to 
the  inadequate  coverage  of  that  index.  The  annual  index  of  the  Service 
gives  120  for  1934  and  126.3  for  1935,  which  is  substantially  above  the 
previous  maximum  of  the  postwar  period  (1929: 115.8).  The  completion 
of  the  preliminary  reports  of  the  1935  census  of  production  supplies  us 
with  confirmatory  evidence.  Data  that  may  be  said  to  present  the 
balance  sheet  of  the  great  depression  have  been  analyzed  by  the  Service 
and  warrant  the  conclusion  "that  physical  output  in  1935  was  about  20 
per  cent  greater  than  in  1930. l  Output  per  operative  employed  in  All 
Industries  increased  still  more,2  engineering  heading  the  list — then  follow 
nonferrous  metals,  textiles,  paper,  and  building — which  suggests  the 
presence  of  quite  as  strong  a  technological  component  as  we  should 
expect  to  find  at  that  cyclical  juncture. 

Related  to  this  technological  component  is  the  fact  that  part  of  the 
innovations  the  consequences  of  which  worked  themselves  out  during 
the  period,  i.e.,  those  connected  with  the  general  reorientation  of  the 
English  economy,  involved  change  of  location  and,  hence,  migration. 
This  alone  would  account  for  an  amount  of  unemployment  supernormal 
even  for  that  phase  of  the  Kondratieff .3  Moreover,  the  general  factors 

1  Special  Memo.  47,  A.  L.  Bowley,  J.  L.  Schwartz,  and  E.  G.  Rhodes,  Output,  Employ- 
ment and  Wages  in  the  United  Kingdom,  1924,  1930,  1935,  p.  27. 

2  Ibid.,  p.  35. 

8  The  slightest  unfavorable  variation  in  employment  is  by  a  sector  of  the  daily  and 
weekly  press  hailed  as  proof  positive  of  the  failing  of  capitalism  and  used  for  derogatory 
comparison  with  forms  of  organization  of  the  bolshevist  and  fascist  type — fascism  being, 


968  BUSINESS  CYCLES 

which  made  for  supernormal  unemployment  in  the  postwar  period  and 
which  we  will  not  list  again,  must  be  borne  in  mind.  On  the  high  level 
set  by  this  frame,  the  fluctuations  of  unemployment  in  England  were 
" normal"  in  our  sense.  Taking  the  number  of  unemployed  insured 
males  as  an  indicator,  and  recalling  that  in  1930  figures  rose  above  the 
one-million  level  of  the  twenties,  we  find  them  moving  at  or  above  2  mil- 
lions (practically)  throughout  1931  and  at  nearly  2.4  millions  during  1932, 
which  also  in  this  respect  was  the  worst  year  though  prices  were  stable 
enough.  A  sharp  decline  followed  in  1933,  which  again  tapered  off 
through  1934  and  part  of  1935  and  then  went  on  at  an  increased  rate, 
approaching  the  one-million  line  in  the  third  quarter  of  1937.  In  Decem- 
ber 1937  a  hyperseasonal  increase  occurred  and  a  little  above  the  figure 
then  reached  (1.31  millions)  unemployment  remained  through  July  1938. 
There  is,  barring  external  factors,  no  reason  for  expecting  very  great 
improvement  in  the  situation  within  the  next  two  or  three  years.  But 
the  armament  or  other  spending  programs  may  do  something.  Also 
migration  has  got  under  way  and  may  reduce  the  "level"  of  the  shorter 
fluctuations. 

This  behavior  of  unemployment  is  strictly  within  the  cyclical  schema 
and  reveals  little,  if  any,  influence  from  monetary  policies.  The  descent 
was  not  slower  or  smaller,  the  recent  ascent  not  quicker  or  greater  than 
we  should  expect  under  English  conditions,  taking  account  of  the  burdens 
imposed  on  the  economic  organism.  But  the  inequalities  as  between 
areas  and  industries  cannot  be  sufficiently  stressed.1  In  June  1936,  for 
example,  unemployment  percentage  in  London,  the  South-east,  the 

in  this  point,  commended  even  by  writers  not  otherwise  in  sympathy  with  it.  There  is 
point  in  this  comparison,  but  it  is  worth  while  stating  in  what  it  consists.  As  far  as  changes 
in  the  productive  organism  involve  geographical  shifts  of  large  numbers  of  workmen 
families,  no  administrative  device  could  entirely  avoid  the  emergence  of  what  would  still 
be  unemployment  in  the  economic  sense.  Though  "temporary,"  it  might  have  to  last  for 
years,  although  a  vigorous  administration  might  prevent  it  from  showing  by  commanding 
the  workmen  to  do  something  meanwhile.  It  is  true,  however,  that  such  periods  need  not 
last  as  long  as  they  are  likely  to  last  in  a  capitalist  organization.  If  workmen  can  be  ordered 
about  regardless  of  their  will  and  immediately  incur  severe  penalties  in  the  case  of  dis- 
obedience, the  importance  of  this,  as  well  as  of  other  sources  of  unemployment,  will  natu- 
rally be  much  diminished.  No  theories  about  vanishing  investment  opportunities  are 
necessary  to  account  for  that  large  increment  of  unemployment  which  owes  its  existence 
to  the  fact  that  in  modern  capitalism  the  workman  is  a  free  and  very  powerful  citizen. 
Unwillingness  to  admit  this  palpable  truth  is  one  of  the  reasons  why  so  many  of  us  take 
refuge  behind  theories  which  without  that  inhibition  they  would  hardly  think  worth 
discussing. 

1  Cf.  the  highly  instructive  analysis  by  D.  G.  Champernowne,  The  Uneven  Distribution 
of  Unemployment  in  the  United  Kingdom  Review  of  Economic  Studies  for  February,  1988. 
It  must  be  remembered  that  these  percentages  refer  to  all  insured  persons  from  which 
English,  unlike  German,  legislation  excludes  domestic  servants. 


THE  WOULD  CRISIS  AND  AFTER  969 

South-west,  and  the  Midlands  was  7.3  per  cent,  while  in  the  North-east, 
North-west,  Scotland,  Wales,  and  North  Ireland  it  was  18.7.  In  August 
1936  the  unemployment  percentage  in  all  industries  was  12. 1;1  but  in 
tramway  and  omnibus  service  for  example,  it  was  only  2.9  and  in  ship- 
building and  port  transport  30.5.2  In  electrical  engineering  it  was  2.6 
per  cent  in  the  first  of  the  two  areas  mentioned,  and  6.3  in  the  second. 
These  examples  suffice  to  show  the  range  of  differences,  which  conclu- 
sively proves  that  much  of  the  total  volume  of  unemployment  was 
intimately  linked  to  conditions  peculiar  to  individual  industries  and  that 
a  great  part  of  the  phenomenon  must  be  missed  by  aggregative  theories, 
diagnoses,  and  remedies.  This  does  not  mean,  of  course,  that  "causes" 
are  not  amenable  to  general  formulation.  On  the  contrary,  these  differ- 
ences are  the  very  results  of  our  process  of  reorganization  and  elimina- 
tion, which  we  designate  by  the  term  competing -down,  including  therein 
the  geographical  rearrangement.  We  are  faced  both  with  states  of 
disequilibrium  and  with  a  movement  toward  a  new  equilibrium.  Hence, 
however  serious  the  temporary  difficulties  and  hardships,  the  problem  is, 
as  far  as  that  goes  and  barring  the  probably  permanent  social  pressure 
exerted  on  the  capitalist  machine,  not  in  itself  a  permanent  one.  It  has 
nothing  to  do  with  structural  peculiarities  of  capitalism — other  than  the 
one  mentioned  in  a  previous  footnote — and,  in  particular,  nothing  with 
any  inherent  inability  of  the  capitalist  mechanism  to  attain  equilibrium 
or  with  any  inherent  tendency  of  it  to  establish  subnormal  equilibria. 

The  wholesale  price  index  of  the  Board  of  Trade  rose,  after  its  relapse 
in  1932,  considerably  in  1933  and  then  remained,  first  because  of  opposite 
movements  in  prices  of  foods  and  nonfoods,  then  because  of  all-round 
stability,  about  at  the  level  attained  until  the  third  quarter  of  1935,  when, 
agreeable  to  expectation,  a  general  rise  set  in.  This  rise,  in  which  prices 
of  materials  were  the  dominant  component,  lasted  to  July  1937 — 
exactly  as  long  as  it  should  have  lasted  according  to  the  cyclical  schema — 
and  for  that  month  carried  the  index  to  80.2  per  cent  of  the  1924  average, 
i.e.,  near  the  1929  average  (82.2).  Cost  of  living  (index  of  Ministry  of 
Labour  corrected  for  seasonals)  stayed  longer  on  the  level  reached  by  the 
relapse  of  1932  and  did  not  rise  materially  until  the  spring  of  1935. 
Then  it  attained  a  maximum  of  92  per  cent  of  the  1924  average  in  April 
1938  and  thus  also  came  near  to  the  1929  average,  which  was  94.  Though 
regular  as  to  fluctuations,  this  behavior  both  of  wholesale  prices  and  of 

1  It  fell  to  about  10  per  cent  in  September  1987,  at  the  high- water  mark  of  prosperity 
and  with  acute  shortage  of  especially  skilled  labor  in  some  lines  and  places.     And  this 
10  per  cent  covers  5.8  for  London  and  the  South-east  area  and  16  per  cent  for  the  Northern 
area,  Wales  and  North  Ireland  displaying  respectively  19.6  and  21. 

2  C/.  Sir  W.  Beveridge,  An  Analysis  of  Unemployment,  Economica  for  November  1986 
and  February  and  May  1937. 


970  BUSINESS  CYCLES 

cost  of  living  is  not  quite  what  we  should  have  expected  from  our  model. 
Recovery  and  prosperity  should  have  done  little  more  than  put  a  tempo- 
rary stop  to  the  fall,  and  price-raising  policies  all  over  the  world  must 
be  held  responsible  for  the  rest.  So  far  as  the  rising  level  was  due  to  the 
increase  in  the  prices  of  imported  raw  materials,  it  obviously  cannot  have 
contributed  much  to  English  recovery,  although  it  may  have  stimulated 
exports  indirectly.  The  terms  of  trade  ceased  in  1936  to  be  as  favorable 
as  they  had  been  during  the  six  preceding  years. 

Pressing  the  new  Index  of  Average  Weekly  Wages  compiled  by  the 
London  and  Cambridge  Economic  Service,  into  a  role  which,  as  we 
know,  it  is  not  quite  qualified  to  play,  we  will  note  that  money  wage 
rates,  which  had  been  very  stable  at  the  1924  average  until  the  middle 
of  1928,  J/£  of  1  per  cent  below  it  to  January  1930  and  a  little  over  98  per- 
cent of  it  until  January  1931,  went  on  slowly  falling  through  1932, 
reaching  94.5  for  the  last  quarter  of  that  year.  Then  weekly  wages  began 
to  rise,  continuing  to  do  so  for  the  rest  of  recovery  and  through  prosperity. 
The  maximum  of  103.25  was  not,  however,  reached  until  April  1938, 
i.e.,  after  recession  had  set  in.  They  have  remained  at  that  level  until 
the  time  of  writing.  This,  as  compared  with  1924,  makes  a  gain  in  real 
rates  of  about  15  per  cent,  or  about  6  per  cent  as  compared  with  1929. 
Thus  the  new  policies  reduced,  but  did  not  eliminate,  the  increase  in  real 
rates  which  at  that  juncture  our  process  would  normally  produce. 

Total  money  wage  bill  (including  salaries)1  was  at  a  minimum  for 
1932  (2,223  million  pounds,  as  compared  with  2,251  million  pounds  in 
1931  and  2,430  million  pounds  in  the  maximum  year  of  the  twenties, 
1929)  and  not  much  higher  in  1933  (2,269  million  pounds)  but  both  years 
were  above  1929  as  to  real  wage  bill — the  total  real  income  of  wage  and 
salary  receivers  thus  behaving  in  a  way  which  again  negatives  the  idea 
of  unrelieved  gloom.  The  figures  for  1934  and  1935  (2,364  and  2,457 
millions)  answer  well  to  our  idea  of  the  behavior  of  money  wage  bills  in  a 
normal  recovery.  The  astounding  steadiness  of  the  English  develop- 
ments and  the  complete  absence  of  "inflation" — so  impressively  asso- 
ciated with  the  mildness  of  the  downturn,  1937-1938 — is  still  more 
obviously  reflected  in  the  behavior  of  total  monetary  home-produced 
income  minus  government  expenditure  (the  maximum  in  the  twenties, 
3,592  millions,  occurred  in  1925  as  the  cyclical  schema  would  have  led  us 
to  expect),  which  was  at  its  cyclical  minimum  of  3,138  (1929:  3,553) 
millions  in  1932  and  recovered  to  3,745  millions  in  1935.  London  country 
and  provincial  clearings  tell  the  same  tale,  which  is  confirmed  by  new 
corporate  capital  issues  for  domestic  purposes — which  did  not  really 
revive  until  1935 — and  by  the  course  of  prices  of  industrial  shares,  which 

1  These  as  well  as  national  income  figures  are  taken  from  C.  Clark,  op.  cit.,  Table  39, 
on  p.  94. 


THE  WORLD  CRISIS  AND  AFTER  971 

rose  practically  steadily  from  the  beginning  of  the  third  quarter  of  1932 
to  the  end  of  the  third  quarter  of  1937. 

E.  The  State-directed  Economy  of  Germany  raises  questions  of 
economic  and  sociological  principle  which  could  be  treated  only  within  a 
research  program  much  wider  and  more  detailed  than  ours.  The  little 
sketch  which  it  is,  nevertheless,  necessary  to  present  will  be  carried  to 
the  spring  of  1938. 

The  outstanding  feature  is  the  rapid  progress,  practically  without 
relapse,  toward  full  employment  of  resources  in  general  and  labor  in 
particular,  in  fact  toward  more  than  that:  unmistakeable  symptoms  of 
overemployment  in  our  sense  show  in  some  industries  before,  in  many 
industries  in,  1937 — among  them  measures  to  relieve  shortage  of  labor.1 
The  Berlin  Institute's  index  of  total  industrial  production  (1928  =  100; 
1929  =  101.4;  1932  =  54)  rose  in  the  monthly  average  to  95.3  for  1935, 
107.8  for  1936,  118.8  for  1937,  120.7  for  the  first  quarter  of  1938,  125.9 
for  April  1938  (April  1937  =  118.5),  the  index  of  what  the  Institute 
classes  as  industrial  producers'  goods,  from  little  over  40  at  the  beginning 
of  1932  to  135  in  April  1938  (equipment  goods  alone,  from  about  30  to 
138),  that  of  industrial  consumers'  goods,  from  less  than  70  to  110.  The 
production  of  raw  steel  rose  to  almost  20  million  tons  in  1937  (United 

1  The  significance  of  what  in  our  terminology  is  overemployment  of  labor  (employment 
greater  than  neighborhood  employment)  has  been  questioned  on  two  grounds.  First,  it 
has  been  pointed  out  that  German  statistics  include  in  the  number  of  employed,  persons  who 
in  the  United  States  would  be  counted  as  unemployed.  Absolute  figures  of  employment 
and  still  more  of  unemployment  are,  in  fact,  internationally  incomparable.  Although 
comparison  of  changes  is  less  misleading,  we  recognize  this  by  stressing  the  obvious  symp- 
toms of  the  presence  of  a  problem  of  shortage.  If  full  employment  were  merely  statistical, 
there  would  obviously  be  no  endeavors  to  cope  with  shortage  (considering  the  level  of  wages 
which  precludes  the  possibility  that  it  is  cheap  labor  that  is  lacking)  by  allowing  temporary 
immigration,  increasing  hours,  increasing  the  employment  of  women,  reinserting  older 
workmen,  and  the  like,  and  no  measures  would  be  taken  about  it,  such  as  the  decree  of 
June  1938,  introducing  general  compulsory  service  for  "tasks  of  particular  public  impor- 
tance." The  statistical  decline  of  unemployment  from  5.7  millions  in  April  1932,  to  0.4  in 
April  1938,  hence,  cannot  be  dismissed  on  this  ground.  At  the  end  of  May  1938,  195,000 
industrial  standard  jobs  were  vacant. 

Second,  it  is  of  course  true  that  the  rebuilding  of  the  army,  the  party  troops  and  party 
activities  in  general,  working  and  other  camps  and  similar  organizations  for  the  young, 
and  workmen's  holidays  absorb  a  great  many  people.  But  some  who  are  thus  absorbed  do 
work  which  would  have  to  be  done  in  any  case,  such  as  traffic  regulation  and  even  industrial 
work.  Moreover,  the  increase  in  the  employable  population  must  be  set  against  that 
absorption,  as  well  as  the  increase  in  women's  and  foreigners'  employment  and  the  higher 
retiring  age.  Finally,  the  increase  in  the  total  number  employed,  from  April  1933  to  April 
1938,  was  (insured  persons  including  servants  and  salaried  employees)  about  7  millions, 
that  is,  greater  by  about  2  millions  than  the  decrease  in  the  number  of  unemployed.  The 
amount  of  invisible  unemployment  in  1933,  though  considerable,  cannot  have  been  nearly 
so  much  as  that. 


972  BUSINESS  CYCLES 

States:  51.7)  and  surpassed  this  record  in  the  first  quarter  of  1938,  reach- 
ing an  all-time  peak  in  spite  of  the  shrinkage  in  exports.  Domestic 
orders  received  by  the  machine  industry  for  the  first  months  of  1938  were 
almost  exactly  seven  times  the  monthly  average  for  1932,  motorcars 
produced  (passenger  and  trucks)  over  six  times.  Electric  power  produc- 
tion nearly  doubled  between  1933  and  1937.  Value  (current  marks)  of 
industrial  and  commercial  construction1  (1932:  0.6  billions)  did  not  rise 
at  all  in  1933,  moderately  in  1934  and  1935,  and  significantly  only  in  1936 
and  1937  (to  1.8  in  the  latter  year),  and  the  value  of  dwelling-house 
construction  (1932:  0.8)  percentually  still  less  (to  2  billions  in  1937). 
If,  nevertheless,  the  building  industry  worked  at  full  capacity  from  1937 
and  efforts  had  to  be  made  to  economize  labor  and  materials,  this  was 
due  to  public  building  (1932:  0.9  billions)  which  increased  by  leaps  and 
bounds  to  6.2  billions  in  1937.  Total  physical  volume  was  then  con- 
siderably above — that  of  dwelling-house  building  about  at — the  level  of 
1928  and  1929,  though  it  is  difficult  to  say  exactly  by  how  much,  because 
of  the  change  in  the  combination  of  cost  factors  that  has  occurred.  The 
Institute's  index  of  building  costs  rose  from  1933  on  but  in  1937  still 
stood  at  only  76.8  per  cent  of  1928. 

This  is  exactly  the  kind  of  performance  that  our  model  would  have 
led  us  to  expect  from  unfettered  capitalism.  Very  obviously,  however, 
capitalism  was  not  unfettered.  In  trying  to  diagnose  the  nature  and 
effects  of  government  leadership  and  control,  we  may  discard  the  policies 
which  revolutionized  the  structure  and  organization  of  the  agrarian  sector: 
vastly  important  though  they  are  from  other  standpoints — in  some 
respects,  perhaps  of  seminal  importance  for  the  solution  of  the  farmer's 
problem — all  that  need  be  noticed  for  our  purpose  is  that  the  index  of 
agricultural  production  was  in  the  average  of  1934-1935  to  1936-1937 
higher  by  19  per  cent  than  in  the  average  of  1927-1928  and  1928-1929, 
and  that  the  index  of  agrarian  prices  in  1936  and  again  in  1937  was  about 
32  per  cent  above  its  minimum  at  the  beginning  of  1933. 

We  may  also  dismiss  the  policy  of  forcing  exports  by  means  of  arrange- 
ments about  blocked  marks,  a  direct  subsidy,  and  bilateral  agreements. 
This  policy,  which  helped  to  increase  values  (current  marks)  by  over 
40  per  cent  from  1934  (the  annual  minimum)  to  1937,  was  dictated 
by  the  necessity  of  providing  exchange  for  the  improvement  of  the 
foreign  debt  situation  and  the  purchase  of  raw  materials.  But  it  counter- 

1The  data  about  values  of  construction  are  taken  from  the  semiannual  surveys 
(Deutschlands  wirtschaftliche  Entwicklung)  published  by  the  Reichs-Kredit-Gesellschaft. 
These  surveys,  interpretations  mainly  based  on  data  of  the  Konjunkturinstitut  and  the 
Statistische  Reichsamt,  may  be  recommended  to  the  reader  as  an  introduction  to  German 
material. 


THE  WORLD  CRISIS  AND  AFTER  973 

acted  the  effects  of  foreign  devaluations  only  in  part,  and  was  not  greatly 
income  generating,  since  the  direct  subsidy  was  financed  by  a  levy  on 
industry. 

More  important  for  our  subject  are  the  efforts  directed  toward  aut- 
arky, under  the  influence  of  which  values  of  imports,  which  had  recovered 
somewhat  in  the  second  half  of  1932,  soon  receded  again  and  then  fluctu- 
ated about  a  horizontal  line  until  their  rise  during  1937.  This  policy, 
while  its  effect  on  welfare  was,  of  course,  negative,  was  an  important 
factor  in  stimulating  prosperity.  A  large  part  of  the  new  investments  in 
industry  was  for  the  development  of  resources  that  were  to  replace 
imported  materials,  for  example,  for  the  development  of  the  iron-ore  or 
the  aluminum  supply,  of  the  production  of  synthetic  gasoline  or  of  the 
staple  fiber  or  of  synthetic  rubber  (Buna).  These  instances  illustrate 
our  distinction  between  conditioning  and  carrying  out  innovation.  Had 
the  response  to  that  policy  consisted  merely  in  an  expansion  on  existing 
lines,  in  an  extension  of,  say,  flax  or  wool  production,  this  would  have 
been  "passive  adaptation"  and  that  policy  would  have  constituted  com- 
plete explanation.  But  that  was  not  all.  New  things  were  done  involving 
the  distinct  entrepreneurial  act  that  constitutes  "creative  adapta- 
tion." In  offering  an  opportunity  for  entrepreneurial  activity  in  pre- 
cisely those  fields,  that  policy  became  no  doubt  responsible  for  many  of 
the  innovations  that  carried  the  prosperity  which  may  be  dated  from  the 
first  quarter  of  1935,  but  only  in  the  sense  implied  in  offering  an  oppor- 
tunity (or  providing  a  condition)  with  which  an  infinite  number  of 
responses  would  have  been  compatible,1  In  appraising  effects  it  is, 
however,  not  only  necessary  to  take  account  of  the  damage  done  to  other 
parts  of  the  industrial  organism — which  it  is  difficult  to  do — but  also  to 
remember  that  it  does  not  follow  that  there  would  have  been  less  entre- 
preneurial activity  in  the  absence  of  the  encouragement  to  press  forward 
in  this  particular  direction.  Entrepreneurs  might  simply  have  pressed 
forward  in  others.  The  reader  should  observe  that  the  failure  to  see  this 
simple  truth  or,  as  we  may  also  put  it,  the  habit  of  exalting  the  importance 
of  the  particular  opportunity  that  has  actually  been  exploited,  at  the 
expense  of  the  role  of  the  opportunity-exploiting  force  or  agent,  is  an  old 

1  By  stating  that  the  policy  of  autarky,  as  such,  conditioned  but  did  not  more  than 
condition  a  certain  type  of  innovation,  we  do  not  mean  that  the  government  did  not  do 
more  than  that.  It  gave  leads.  It  exerted  pressure.  It  helped  in  various  ways  in  financ- 
ing and  promoting.  Production  of  synthetic  gasoline  was,  for  example,  subsidized  by  a 
levy  on  the  brown-coal  industry.  For  the  expansion  of  iron-ore  production,  the  Reichs- 
werke  Hermann  Goring  were  founded,  of  which  the  Reich  took  most  of  the  common  stock. 
And  there  were  many  cases  of  pure  state  enterprise.  This  active  leadership  was,  of  course, 
something  very  different  from  mere  "control"  or  "regulation"  and  also  from  mere  condi- 
tioning. But  it  must  be  distinguished  from  the  policy  of  autarky  as  such. 


974  BUSINESS  CYCLES 

error  which  we  have  met  over  and  over  again  on  our  way  (for  example,  in 
the  discussions  (Chap.  VI)  about  the  "rise  of  capitalism")  and  which  is 
a  fertile  source  of  faulty  diagnosis. 

In  a  different  form,  the  same  error  must  be  guarded  against  in  a 
discussion  of  the  conspicuous  success  of  the  spending  policies  of  the  Ger- 
man government.  We  must  strictly  distinguish  between  two  types:  one 
of  them  was  simply  of  the  pump-priming  kind — the  German  words 
being,  as  before,  Ankurbelung  and  Arbeitsbeschaffung — in  the  sense  that 
it,  and  other  expenditure  induced  by  it,  was  additive  to  that  system 
expenditure  which  would  have  obtained  in  its  absence,  while  the  other 
was  substitutive  in  the  sense  that  it  took  the  place  of  part  of  that  system 
expenditure.  This  distinction  does  not  coincide  with  that  between 
income-generating  expenditure  and  non-income-generating  expenditure, 
for  public  expenditure  may  be  strongly  income  generating  and  yet  need 
not  be  additive  in  that  sense:  it  would  not,  if  income-generating  private 
investment  were,  for  example,  at  the  same  time  curtailed  by  the  same 
amount.  We  cannot,  of  course,  draw  the  line  according  to  the  objects 
for  which  given  income-generating  outlays  were  made.  Roads,  canals, 
public  buildings,  beautification  of  cities,  armaments  may  all  come  into 
either  category,  although  we  will  call  the  second  category  "armaments/* 
a  potiori,  in  spite  of  the  fact  that  such  things  as  the  Rhine-Danube  Canal 
(act  of  May  11,  1938)  also  enter  into  it.  But  we  can  pretty  definitely 
draw  a  line  in  time :  income-generating  expenditure,  whatever  its  motive, 
was  primarily  pump  priming  or  additive  until  about  the  first  quarter  of 
1935  and  primarily  substitutive  after  that. 

Pump-priming  income  generation  was,  though  only  an  enlarged  edi- 
tion of  previous  policies,  part  of  the  so-called  First  Four-year  Plan  and 
of  a  recovery  policy  which  shows  much  family  likeness  to  that  of  the 
United  States.  What  might  be  termed  the  German  AAA  met  a  different 
situation  in  a  different  spirit,  but  is  similar  to  its  American  counterpart  in 
its  aim  of  raising  agricultural  prices.  The  German  NRA  was  the  act  of 
July  15,  1933  (introducing  compulsory  cartels).  It  stressed  restriction 
of  production  and  of  real  investment — by  order  of  the  Reichswirtschafts- 
ministerium — still  more  than  its  American  counterpart.  Founding 
new  firms,  constructing  new  plants  and  resuming  operation  of  plants 
that  had  been  shut  down,  installing  new  machinery  or  otherwise  increas- 
ing capacity  were  for  definite  periods  prohibited  in  a  long  list  of  industries. 
Later  on,  these  cartels  were  in  other  respects  tightened  and,  as  far  as  not 
superseded  by  other  controlling  agencies — the  commissioner  for  the  super- 
vision of  pricing  in  particular;  decree  of  Nov.  5,  1934,  which  revived  an 
office  already  created  Dec.  8,  1931 — made  a  permanent  instrument  of 
government  control  over  production.  But  as  a  part  of  recovery  policy, 
prohibition  to  invest,  except  in  special  cases  (e.g.,  prohibition  of  opening 


THE  WORLD  CRISIS  AND  AFTER  975 

new  retail  shops),  was  discontinued  as  soon  as  it  was  thought  that  it 
had  served  its  recovery  purpose  and  replaced  by  compulsion  to  invest 
according  to  government  order.  The  first  case  (October  1934)  thus 
taken  in  hand  was  that  of  the  brown-coal  industry  mentioned  in  the  last 
footnote.  It  was  organized  into  a  compulsory  "community"  (Gemein- 
schaft)  and  directed  to  make  a  levy  on  its  members  with  which  to  finance 
the  Braunkohle-Benzin  A.G.  for  the  purpose  of  developing  hydrogenation.1 
Within  this  arsenal  of  recovery  measures  pump  priming  went  on  by 
means  of  relief  and  public  works  expenditure,  as  elsewhere  largely 
financed  by  short  government  paper  discounted  at  the  Reichsbank  or 
taken  under  rediscount  promise  of  the  Reichsbank  by  banks,  saving 
banks,  and  other  institutions.  It  is  not  easy  to  estimate  its  amount,  but 
excluding  the  investment  done  by  the  state  railways  and  the  post  office 
the  writer  believes  it  to  have  been,  until  the  middle  of  1935,  of  the  order  of 
magnitude  of  from  3  to  4  billion  marks.2  Previous  argument  justifies 
the  statement  that  this  expenditure,  coming  as  did  the  American  spending 
program  of  1933,  after  the  event,  i.e.,  after  the  lower  turning  point  of 
the  cycle,  did  not  break  the  spell  of  a  depression  that  would  otherwise 
have  kept  the  system  indefinitely  at  the  minimum  point.  But  its  suc- 
cess in  helping  and  accelerating  recovery  was  striking,  as  shown  by  the 
unbroken  improvement  and  the  rapid  rate  of  absorption  of  unemploy- 
ment. Since  such  success  does  not  attend  all  pump  priming,  it  is  reason- 
able to  attribute  it  to  the  manner  in  which  it  was  done  in  this  case  and  to 
concomitant  policies.  The  sums  disbursed  were  comparatively  moder- 
ate. They  were  expended  with  strict  economy.  Creation  of  purchasing 
power  was  an  incident,  but  it  was  not  pursued  as  an  end.  Speculation 
was  not  encouraged.  Infraction  of  social  discipline  was  discouraged. 
No  attempt  was  made  to  raise  costs.  Monetary  wage  rates,  in  particular, 
were  regulated  with  a  view  to  stability  at  a  level  not  much  above  depres- 
sion minimum  (see  below).  Saving  and  accumulation  were  encouraged, 
or  as  little  discouraged  as  possible,  and  in  many  instances  enforced.3 

1  As  we  shall  see  below,  compulsion  to  invest  in  some  lines  frequently  implied  prohibi- 
tion to  invest  in  other  lines;  but  these  prohibitions  were  no  longer  dictated  by  the  recovery 
purpose  and  carry  a  different  meaning.     There  was  a  third  class,  viz.,  the    restrictions 
dictated  by  the  raw-material  and  exchange  situation.     These,  however,  do   not  interest 
us  here. 

2  The    three    programs    for    providing    employment     (Arbeitsbeschaffungsprogramme) 
alone  amounted  to  1,900  million  marks.     The  writer  has  added  the  whole  of  the  government 
subsidies  for  repairing  buildings  (500  millions)  and  a  highly  conjectural  part  of  the  expendi- 
ture on  motor  roads  and  other  items  of  public  construction  not  included  in  those  1,900 
millions.     The  chief  difficulty  is  in  determining  how  much  of  the  total  created  additional 
income. 

3  Space  forbids  our  entering  into  technique,  but  the  main  headings  must  be  mentioned. 
First,  that  policy  of  wages  was  on  balance  conducive  to  increasing  the  sum  total  of  savings. 
Second,  taxation  was  at  first  somewhat  readjusted  in  a  way  favoring  savings  and  later  on, 


976  BUSINESS  CYCLES 

And  all  this  minimized  dislocating  and  maximized  precisely  that  part  of 
stimulating  effects  which  does  not  produce  relapses.  The  strength  of  the 
" fascist"  state  as  against  group  interests,  and  its  fundamental  attitude 
to  economic  life — which  for  it  is  not  an  end  in  itself  but  a  subordinate 
servant  of  extra-economic  national  goals — in  this  case  facilitated  a 
behavior  in  accordance  with  the  rules  of  long-run  economic  rationality. 
Government  expenditure  was  not  less  conspicuous  and  government 
leadership  was  more  so  in  the  subsequent  prosperity  than  it  had  been 
during  the  recovery.  But  it  thenceforth  acquired  a  different  economic 
character.  We  speak  of  armament  expenditure  as  belonging  to  a  differ- 
ent category,  neither  because  it  was  evidently  not  motivated  as  a  pump- 
priming  measure — that  would  not  matter — nor  because  it  was  not 
income  generating — for  it  was — but  because  it  was  not  additive  in  the 
sense  defined  above.  There  is,  of  course,  no  reason  to  believe  that 
results  would  have  been  any  different  if  the  government's  demand  had 
been  for  washing  machines  and  baby  carriages  instead  of  for  war  equip- 
ment. But  there  is  also  no  reason — other  than  an  ad  hoc  assumption 
based  on  prejudice — for  believing  that,  once  started  upon  the  road  toward 
full  employment,  the  system  would  have  failed  to  reach  it  and  to  embark 
upon  a  prosperity  phase  if  there  had  been  no  such  government  demand  at 
all.  For,  on  the  one  hand,  an  equal  amount  of  existing  monetary  means 
and  of  facilities  for  creating  additional  ones  would  have  been  released  to 
firms  and  households,  and,  on  the  other  hand,  private  investment  pro- 
grams abounded  and  households  were  obviously  not  averse  to  expanding 
private  consumption.  That  this  was  so  is  proved  by  the  fact  that,  in 
order  to  finance,  and  to  provide  the  physical  resources  for,  its  own  pro- 
gram of  investment  and  consumption — guns  are  consumers*  goods — the 
government  had  to  place  severe  restrictions  on  private  issues,  real  invest- 
ment, and  consumption  alike.  In  particular,  the  restrictions  on  issues 
and  investments  other  than  those  within  the  government  program  can 
no  longer  have  been  motivated  by  a  wish  to  overcome  depression.  Hence, 

when  the  burden  had  to  be  increased  (e.g.  by  the  increase  in  the  corporation  tax  from  20  to 
30  per  cent,  August  1986),  it  was  done  in  a  way  which  did  at  least  not  differentiate  against 
saving.  Third,  by  an  act  regulating  the  distribution  of  profits  of  corporate  enterprise 
(Dec.  4,  1934),  limitations  were  imposed  not  on  undivided  but  on  distributed  profits. 
Fourth,  where  real  investment  in  industrial  plant  and  equipment  was  made  compulsory, 
this  was  done  in  such  a  way  as  to  make  saving  or  accumulation  practically  compulsory  too. 
Fifth,  interest,  while  regulated,  was  regulated  at  a  comparatively  high  level.  This  may 
have  had  something  to  do  with  the  increase  in  saving  deposits — though  the  reader  knows 
how  little  faith  the  writer  places  in  any  figures  about  saving — which  (annual  increase  in 
saving  banks  and  other  institutions)  was  as  follows:  620  million  marks  in  1933;  957  in  1934; 
1,244  in  1935;  1,084  in  1936;  1,823  in  1937.  The  Reichsamt  estimates  (this,  however, 
is  still  more  doubtful)  the  total  amount  of  monetary  savings  at  1,571  millions  for  1933  and 
at  6,895  for  1937. 


THE  WORLD  CRISIS  AND  AFTER  977 

they  also,  like  government  expenditure,  acquired  a  new  meaning:  they 
were  continued  in  order  to  reserve,  for  the  government,  factors  of  pro- 
duction which  were  known  to  be  on  their  way  to  other  employments  and 
had  to  be  deflected  from  them.  But  this  implies  that  government  did 
not  create  a  demand  for  those  factors  which  would  otherwise  have  been 
lacking,  but  that  it  only  substituted  its  own  demand  for  that  which  would 
otherwise  have  been  forthcoming  from  other  sources — which  is  what  we 
wished  to  prove.1 

The  argument  does  not  deny  that  the  presence  of  so  obvious,  strong, 
and  steady  a  demand  greatly  facilitated  matters,  and  still  less  that  the 
government's  labor  and  saving  policies  substantially  contributed  to  the 
success  revealed  by  the  output  and  employment  figures.  Nor  does  it 
deny  the  importance  of  the  momentous  change  involved  in  the  fact  of 
actual  management  of  industry  by  the  state.  But  it  still  follows  that 
the  prosperity  was  only  state-directed  and  not  state-created  and  that  it, 
hence,  fits  very  much  more  closely  into  our  schema  than  one  would 
believe  at  first  sight.  The  developments  to  1937,  in  fact,  make  perfectly 
good  Juglar  recovery  and  prosperity  phases.  The  interesting  question  as 
to  what  inference  that  suggests  about  future  cyclical  phases  in  the  Ger- 
man economy  and,  in  particular,  depressive  ones  is  perhaps  not  difficult 
to  answer.  Theoretically  it  is  possible  so  to  plan  the  sequence  of  innova- 
tions as  to  iron  out  cycles;  but  after  supernormally  strenuous  periods  of 
advance  there  will  be  recessions  in  our  sense  even  in  the  corporative  state ; 
most  of  the  symptoms  of  depression,  however,  need  not  occur  at  all  or 
can  be  made  to  disappear  promptly  by  so  powerful  a  central  authority. 

It  remains  to  survey  in  the  light  of  the  above  analysis  the  behavior 
of  a  few  additional  series.  Prices  were  strictly  regulated  and,  owing  to 
the  practically  complete  socialization  of  Germany's  international  eco- 
nomic relations,  divorced  from  foreign  levels  and  deprived  of  any  influ- 
ence on  international  movements  of  commodities.  Moreover,  they  were 
also  subject  to  all  sorts  of  distortions  resulting  from  the  German  situation. 
But  they  did  not  on  that  account  entirely  fail  to  conform  to  the  logic  of 
our  model.  This  is  due  to  the  fact  that  regulation  asp  racticed  not  only 

1  For  the  obviousness  of  that  proof  we  are  indebted  to  the  government's  determination 
to  do  the  thing  in  as  uninflationary  a  way  as  possible.  If  instead  of — practically — com- 
mandeering the  productive  resources  of  the  nation,  it  had  bid  for  them  in  free  markets 
(which  could  have  been  done  only  by  means  of  much  additional  credit  creation),  its  demand 
would  have  been  additive,  and  it  would  have  been  more  difficult  to  show  that  in  real 
terms  our  diagnosis  would  have  held  true  even  then.  It  should  be  observed  that  while 
a  much  more  general  (and  also  more  correct)  theoretical  argument  could  be  produced  in 
support  of  the  economic  truth  we  strive  to  grasp,  our  proof  makes  use  only  of  the  particular 
conditions  of  present-day  Germany  and,  hence,  cannot  itself  be  generalized.  Nor  does  it 
prove  that  the  intensity  of  prosperity  under  state  leadership  was  exactly  what  it  would 
have  been  without  it.  It  might  have  been  less  intense.  It  might  have  been  more  so. 


978  BUSINESS  CYCLES 

caused  but  also  avoided  some  of  those  deviations  which  under  modern 
conditions  of  pricing  we  are  likely  to  meet.  During  the  five  years  that 
have  elapsed  since  the  government  came  into  power,  the  index  of  whole- 
sale prices  rose  (April  1933  to  April  1938)  from  64.8  (1928  =  100)  to  75.4. 
But  this  rise  was  due  mainly  to  the  prices  of  agricultural  products  and  of 
imported  materials — rubber  rose  by  over  600  per  cent.  Finished  indus- 
trial producers'  goods  fell  somewhat,  while  finished  industrial  consumers' 
goods  rose  by  almost  25  per  cent.  The  official  index  of  cost  of  living  rose 
during  the  same  time  by  8.5  per  cent.1  The  picture  thus  reflects  the 
tendency  appropriate  to  the  Kondratieff  phase  if  we  take  account  of  a 
number  of  circumstances  peculiar  to  Germany.  It  does  not  reveal  any 
intention  of  government  to  raise  the  price  level  as  such.  Excepting  the 
agrarian  sector,  there  seems  in  fact  to  have  been  no  tendency  to  use 
increase  in  prices  as  a  stimulant.  On  the  contrary,  efforts  were  made  to 
prevent  the  expansion  of  the  circulating  medium  from  having  that  effect. 
Nevertheless,  and  in  spite  of  the  continuing — though  perhaps  decreasing 
— losses  incident  to  the  policy  of  forcing  exports,  of  increases  in  operating 
costs  due  to  increasing  prices  of  raw  materials  or  to  the  unsatisfactory 
nature  of  substitutes  or  to  working  beyond  optimum  capacity,  and  of 
increases  in  total  cost  due  to  the  necessity  of  liberal  allowances  for 
depreciation,  industrial  net  earnings  increased  through  all  the  years 
under  survey.  But  they  did  so  at  a  decreasing  rate  suggestive  of  the 
approach  of  recession.  So  did  stock  prices. 

That  price  policy  was  made  possible  by  the  government's  wage  policy, 
which  up  to  the  point  of  full  employment  produced  the  very  results  that 
we  would  expect  under  conditions  of  perfect  competition.  According  to 
the  official  data,  average  annual  hourly  rates  rose  to  only  76  pfennigs  for 
1937  (from  the  minimum  of  70  in  1933)  roughly  as  cost  of  living2  and  thus 
remained  far  below  the  maximum  of  1929  (96  pfennigs)  in  monetary,  and 
somewhat  below  it  (by  about  3  per  cent)  in  real  terms.  This  policy  of 

1  One  reason  why  cost  of  living  did  not  increase  more  than  that  was  the  unrelenting 
pressure  that  was  exerted  on  retailers'  margins.     In  food  retailing  net  revenue  of  many 
small  shops  is  said  to  have  fallen  to  as  little  as  1,000  marks,  the  equivalent  in  purchasing 
power  of  roughly  $800  to  $400  per  year.     Of  course,  this  was  due  not  only  to  price  policies 
but  also  to  various  restrictions  which  limited  physical  volume  of  sales,  increased  difficulties 
in  purchasing,  and  to  other  causes.     Fundamentally,  there  is  hardly  room  for  the  independ- 
ent retailer  of  the  capitalist  type  in  a  community  that  extends  regulation  to  consumption. 

2  If  welfare  considerations  were  relevant  to  our  subject,  we  should  have  to  take  account, 
on  the  one  hand,  of  the  change  in  the  taxation  of  wages  (3.5  per  cent  to  1929,  then  3  per  cent 
to  1982,  3.5  per  cent  for  1933,  then  4  per  cent  and  4.5  per  cent  for  1936  and  1937;  including 
poll  tax;  data  from  Reichsamt,  Institut  fiir  Konjunkturforschung  and  reports  of  Reichs- 
kreditgesellschaf  t)  though  not  the  contributions  to  social  insurance,  and,  on  the  other  hand, 
of  a  number  of  benefits,  holidays,  cheap  excursions,  marriage  loans,  protection  against 
dismissal,  and  the  like  that  are  difficult  to  evaluate  but  of  considerable  importance. 


THE  WORLD  CRISIS  AND  AFTER  979 

making  and  keeping  labor  a  cheap  factor  of  production  greatly  helped  to 
increase  the  total  income  of  the  working  class.  The  industrial  wage  bill 
just  about  doubled  in  utter  disregard  of  high-wage  theories  between 
1933  and  1937  (1933:  5,921  millions;  1937:  11,900  millions;  the  Saar 
country  being,  for  the  sake  of  comparability,  excluded  in  both  years). 
An  investigation  of  the  Berlin  Institute  states  that  about  65  per  cent  of 
this  increase  is  attributable  to  increase  in  numbers  employed;  about 
11.5  per  cent  to  longer  hours;  about  10  per  cent  to  promotion  to  better 
paid  jobs;  and  about  13.5  per  cent  to  increase  in  rates.  The  sum  total 
of  all  wages  and  salaries  increased  only  from  12.1  billions  in  1933  to  20.9 
billions  in  1937  and  total  national  income  only  from  46.6  to  68.5  billions 
(1929:  75.9),  as  a  consequence  of  a  policy  which,  on  principle,  strove  to 
conserve  rather  than  to  reverse  the  downward  revision  of  monetary 
values  effected  by  the  depression.  Real  per  capita  income  increased  all 
along  and  in  1936  and  1937  surpassed  that  of  1929.  Though  the  results 
of  recovery  and  prosperity  were  to  a  great  extent  absorbed  by  public 
consumption  and  investment,  the  consumption  of  the  masses  expanded 
in  the  field  of  the  commodities  of  modern  life,  household  gadgets  and  the 
like.  It  contracted  in  some  departments,  where  it  ran  up  against  the 
exigencies  of  autarky.  But  per  capita  figures  lend  only  feeble  support 
to  the  widespread  belief  that  there  was  all-round  contraction  in  the 
consumption  of  food.  Comparison  of  figures  for  1937  and  1929  shows 
indeed,  significant  decrease  in  the  cases  of  citrus  fruit,  beer,  eggs,  and 
margarine;  and  an  insignificant  one  in  the  case  of  wheat.  But  per  capita 
consumption  was  higher,  e.g.,  in  the  cases  of  rye,  meat,  fish,  butter,  sugar, 
and  coffee. 

The  management  of  money  and  credit  was  facilitated  and  its  effective- 
ness greatly  increased  by  that  policy  of  wages,  by  the  government's 
attitude  toward  and  perfect  control  over  savings,  and  by  being  completely 
entrusted  to  one  very  able  man.  But  that  part  of  it  which  interests  us 
here1  did  not,  either  in  the  problems  to  be  solved  or  in  the  methods  of 

1  The  rest  of  it,  viz.,  the  management  of  the  relations  of  the  German  to  foreign  monetary 
and  credit  systems — complete  control  of  transactions  in  foreign  exchange  and  all  move- 
ments of  commodities  and  balances  was  only  the  most  obvious  tool  of  that  management — 
also  bears  upon  our  subject,  not  only  through  the  "insulation"  of  the  German  money 
market  which  it  achieved,  but  also  through  its  influence  upon  the  investment  process. 
But  we  cannot  enter  into  it.  It  must  suffice  to  state  that  this  part  of  German  monetary 
policy  was  chalked  out  for  government  and  Reichsbank  by  the  debt  situation  of  1932  and 
the  methods,  stand-still  agreements  and  moratoria,  by  which  it  was  then  handled.  The 
facts  that  certain  extra-economic  policies  of  the  National- Socialist  government  produced  a 
"flight  of  capital"  and  that  foreign  credits  were  both  difficult  to  get  and  from  the  national 
standpoint  undesirable,  of  course  intensified  the  difficulties  of  that  situation  and  of  the 
problems  of  raw-material  supply.  Foreign  devaluations  and  some  repayments  and 
repurchases,  all  of  which  reduced  the  foreign  debt  from  about  19  billion  marks  at  the  begin- 


980  BUSINESS  CYCLES 

solving  them,  fundamentally  differ  from  English  or  American  manage- 
ment. The  German  money  market  being  almost  completely  insulated — 
much  more  so  than  it  could  have  been  insulated  by  a  mere  abandonment 
of  the  gold  parity — the  Reichsbank,  while  allowing  its  holdings  of  gold 
and  exchange  to  dwindle  to  practically  nothing  (a  little  more  than  70 
million  marks  before  the  absorption  of  the  Austrian  National  Bank), 
secured  powers  to  embark  upon  open-market  operations  and  to  "cover" 
her  notes  exclusively  by  domestic  bills  and  certain  types  of  fixed-interest 
securities  through  the  legislation  announced  Oct.  17,  1933.  It  also 
secured  more  stringent  control  over  other  banks  through  the  Banking 
Act  (Gesetz  iiber  Kreditwesen)  of  1934.  Thus  armed,  it  immediately 
set  about  expanding  the  volume  of  credit,  but  only  within  the  require- 
ments of  pump-priming  public  expenditures.  To  the  end  of  1935  it 
increased — together  with  its  affiliate,  the  Gold  Discount  Bank — what 
we  may  term  central  credit  outstanding  (very  roughly  corresponding  to 
federal  reserve  credit  outstanding)  by  about  2.7  billion  marks,  which 
was  sufficient  to  unfreeze  what  in  1933  and  1934  there  still  was  to  unfreeze 
in  the  industrial  and  the  banking  structure  and  to  help  the  market  to 
absorb  the  various  kinds  of  short  government  paper  which  financed  that 
expenditure  (about  1.5  millions  of  tax  refund  bills — Steuergutscheine, 
introduced  by  the  Papen  plan — several  billions  of  bills  for  the  financing 
of  employment — Arbeitsbeschaffungswechsel — other  special  bills — Sonder- 
wechsel — and  ordinary  treasury  bills,  all  of  which  produced  a  net  increase 
in  all  kinds  of  bills  outstanding  of  between  5  and  6  billions,  or  about 
50  per  cent,  from  the  last  quarter  of  1932  to  the  last  quarter  of  1935). 
No  great  pressure  was  brought  to  bear  on  short  rates.  In  fact,  the 
method  of  allowing  that  mass  of  short  paper  to  go  to  the  open  market 
relieved  member  banks  of  idle  funds  and  idle  facilities  and  thus  of  the 
necessity  to  compete  for  other  outlets.  Short  rates  fell  in  the  natural 
course  of  things,  but  remained  comparatively  high  throughout :  bank  rate 

ning  of  1933  to  about  10  billions  at  the  beginning  of  1938,  brought  but  little  relief.  Under 
these  circumstances  a  policy  developed  from  what  at  first  had  been  temporary  emergency 
measures.  It  may  be  described  either  as  an  attempt  to  secure  some  effects  of  devaluation 
without  devaluating  or,  more  tellingly,  as  graduated  devaluation,  the  rate  varying  from 
about  40  per  cent  to  zero.  The  state,  completely  controlling  foreign  economic  relations 
and  therefore  being  in  the  position  of  a  discriminating  monopolist,  discriminated  by 
means  of  a  schedule  of  prices  for  marks  graduated  according  to  the  use  to  be  made  of  every 
one  of  them  and  to  the  elasticity  of  the  demand  in  each  use:  holiday  making  in  Germany, 
for  example,  is  for  the  foreigner  a  highly  substitutable  commodity,  hence  the  offer  of  a 
particularly  cheap  traveling  mark.  It  follows  that  it  is  not  correct  to  say  without  qualifica- 
tion that  Germany  has  not  devaluated  at  all,  but  that  it  is  also  not  correct  to  say  that  the 
official  parity  price  of  the  mark  is  meaningless  and  nothing  but  a  sham.  If  it  were  a  sham, 
the  method  of  export  subsidies,  which  is  highly  distasteful  both  to  government  and  indus- 
try, obviously  need  not  have  been  resorted  to. 


THE  WORLD  CRISIS  AND  AFTER  981 

stood  at  4  per  cent  from  1933  to  1938,  rate  on  prime  bankers*  acceptances 
(Privatdiskont)  fell  only  from  3.88  to  2.88,  daily  money  rate  from  5.11  to 
2.63.  This  need  not  have  been  so.  But  recovery  was  financed  at 
relatively  high  rates  in  order  to  facilitate  normalization,  which  was  kept 
steadily  in  view.  Long  rates  were  in  some  sectors  reduced  by  government 
action,  but  where  they  were  not,  especially  in  the  bond  market,  adapta- 
tion to  the  liquidity  in  the  short  market  was  allowed  to  proceed  but 
slowly.  Prices  of  6  per  cent  industrial  bonds  reached  parity  only  in 
1935 — they  remained  above  it  after  1936,  when  the  rate  was  reduced  to 
5  per  cent — and  the  price  of  4.5  per  cent  governments — until  the  conver- 
sion of  Mar.  31,  1935,  the  rate  was  6  per  cent — did  not  approach  it  until 
May  1938  (99.7). 

"Expansive"  short  financing  continued  during  the  prosperity,  but  it 
was  increasingly  replaced  by  government  issues  offered  for  public  sub- 
scription. These  had  been  insignificant  in  1933  and  1934,  but  amounted 
to  1,636  millions  in  1935;  2,670  in  1936;  3,150  in  1937;  and  to  1,934  in 
the  first  quarter  of  1938,1  when  a  program  of  normalization  was 
announced.  After  one  more  issue  of  the  "expansive"  type — not  redis- 
countable,  however,  at  the  Reichsbank,  though  eligible  as  collateral  for 
loans — financing  by  "  special  bills  "  was  to  be  discontinued;  the  Reichsbank 
was  to  help  in  absorbing  the  existing  ones  but  no  longer  to  rediscount  them 
or  any  new  ones;  the  Reich  was  to  finance  from  taxation  or  ordinary 
loans,  industry  and  trade  from  ordinary  bank  credit.  In  itself  and 
barring  disturbance  by  political  events,  this  program  is  perfectly  sound 
both  in  the  sense  that  it  is  possible  to  carry  it  out  and  that  there  is  no 
reason  why  doing  so  should  produce  a  normalization  crisis.  But  very 
heavy  taxation  is  an  essential  element  of  it.  In  consequence  of  the 
increase  in  incomes  and  in  rates  of  taxes,  the  revenue  of  the  Reich 
increased  by  7.3  billions  between  the  fiscal  years  1932-1933  and  1937- 
1938,  and  there  is,  of  course,  the  decline  in  expenditure  on  unemployment 
relief  amounting  to  nearly  2.5  billions.  But  the  tax  on  corporate  earnings 
contributed  over  1,400  millions  to  that  increase,  and  the  income  tax — 
exclusive  of  the  tax  on  wages — nearly  1,700.  Whether  in  the  fascist  state 
the  effects  of  this  will  in  the  long  run  differ  from  those  we  should  have 
to  expect  in  other  social  conditions,  remains  to  be  seen. 

The  behavior  of  banking  series  reflects  the  processes  of  state-directed 
and  state-financed  recovery  and  prosperity,  and  can  be  easily  interpreted 
from  this  standpoint.2  When  foreign  observers  in  the  second  quarter 

1  Redemption  of  all  kinds  of  public  debts  during  1935  to  1938  amounted  to  1,750 
millions  according  to  the  Reichskreditgesellschaft. 

2  This,  however,  applies  only  to  the  most  general  contours.     As  we  have  seen  before  this, 
German  bank  statistics  are  anything  but  easy  to  analyze  in  detail.     Difficulties  have, 
among  other  things  because  of  the  multiplication  of  intermediate  instances  (Arbeitsbe- 


982  BUSINESS  CYCLES 

of  1933  inferred  from  the  continuing  shrinkage  of  loans  and  deposits — 
by  deposits  we  mean  the  German  Kreditoren  and  not  deposits  in  the 
German  sense — that  there  was  no  genuine  recovery,  they  not  only  over- 
looked that  foreign  devaluations  and  repayments  to  foreign  creditors 
would  automatically  produce  a  shrinkage  of  balance-sheet  items,  but  also 
that  the  domestic  processes  of  unfreezing  and  of  repaying  emergency 
credits  would  naturally  have  the  same  effect.  But  German  observers 
who  did  not  fail  to  urge  this,  still  expected  that  expansion  of  member 
bank  credit  would  presently  set  in.  Consequently  they  were  disap- 
pointed at  what  looked  like  absence  of  "secondary  effects,"  the  failure 
of  public  expenditure  to  stimulate  private  investment,  and  so  on.1  In 
fact,  the  Debitoren  (loans)  of  the  five  great  banks  even  declined  through- 
out prosperity  and  were  but  little  over  3  billion  marks  by  the  end  of  1937, 
after  having  been  a  little  over  4  billions  at  the  end  of  1933  and  about  6 
billions  at  the  end  of  1929. 2  Their  deposits  including  "saving"  deposits, 
over  10  billions  in  1929,  moved  on  a  falling  level  through  1935  and  slowly 
recovered  to  but  6.8  by  April  1938.  Cash  also  declined.3  Bills,  however, 
mostly  of  course  the  "special"  bills,  more  than  doubled  from  the  first 
quarter  of  1933  to  the  last  quarter  of  1937  and  investments — owned 
securities,  eigene  Wertpapiere — also  increased  considerably. 

There  is  nothing  to  wonder  at  in  this  picture.  Its  features  are  the 
consequence  of  government  financing.  Business  did  not  go  to  banks 
for  what  it  received  from  government,  hence  business  expansion  did  not 
directly  show  in  bank  statistics  except  in  money  in  circulation — which 
amounted  to  5,418  at  the  end  of  February  1933,  and  to  7,219  at  the  end 
of  February  1938,  the  last  figure  unaffected  by  the  annexation  of  Austria 

schqffungsbanlcen  and  other)  and  because  of  the  lumping  of  "special"  with  other  bills, 
not  decreased  since  1932.  We  shall  confine  ourselves  to  the  Reichsbank  and  the  big  Berlin 
banks,  but  will  take  this  opportunity  to  mention  two  additional  points.  First,  we  have 
seen  what  plight  the  municipalities  maneuvered  themselves  into  by  their  liberal  expendi- 
ture during  the  twenties,  which  during  the  crisis  issued  into  complete  breakdown  of  munici- 
pal finances.  One  of  the  first  tasks  of  financial  reconstruction  was,  therefore,  the  refinanc- 
ing especially  of  their  short  debts  which  was  done  by  the  Gemeindeumschuldungsgesetz. 
Second,  in  order  to  keep  the  figure  of  Reichsbank  credit  outstanding  as  low  as  possible,  the 
Golddiskontbank  issued  one-name  paper  (Solawechsel)  to  the  banks  in  order  currently  to 
absorb  their  liquid  means  and  to  use  these  for  relieving  the  Reichsbank. 

1  For  some  of  these  comments,  as  well  as  for  Professor  Bresciani-Turroni's  interpreta- 
tion, see  the  latter's  article,  The  Multiplier  in  Practice:  Some  Results  of  Recent  German 
Experience,  Review  of  Economic  Statistics,  May  1938. 

2  The  commodity  loans  (Warenvorschusse) ,  which  at  the  end  of  1929  had  reached  2  bil- 
lions, were  but  a  few  hundred  millions  throughout  1935  to  1937.     We  do  not  emphasize  the 
decline  in  financing  of  international  trade  (Rembours-Kredit)  and  in  stock  exchange  loans 
(Reports)  because  they  are  accounted  for  by  obvious  special  reasons. 

3  Since  a  very  large  part  of  the  other  assets  could  at  a  moment's  notice  be  liquidated  at 
the  Reichsbank,  there  was,  in  fact,  little  object  in  holding  cash  beyond  current  requirements. 


THE  WORLD  CRISIS  AND  AFTER  983 

— and  in  the  bills.1  And  multiplier  effects  and  the  stimulation  of  private 
investment  show  during  recovery  in  the  sphere  of  production  and  employ- 
ment rather  than  in  the  sphere  of  regular  bank  credit,  while  there  is  no 
point  at  all  in  looking  for  the  former  after  full  employment  had  been 
reached.  Those  effects  are,  no  doubt,  difficult  to  evaluate,  but  during 
what  we  have  defined  as  the  pump-priming  period  their  presence  seems 
clear  enough  wherever  they  were  not  obstructed  by  difficulties  about  raw 
materials  and  so  on.  During  what  we  have  called  the  armament  period 
— the  prosperity  phase — the  monetary  processes  of  an  ordinary  prosperity 
were  replaced  by  the  monetary  processes  of  a  state-financed  one,  as  other 
processes  were  by  the  processes  of  "investment"  for  the  purposes  of 
armament  and  of  autarky  (the  Second  Four-year  Plan).  The  tapering 
off  of  pump  priming  thus  becomes  invisible,  because  the  flow  of  prosperity 
expenditure  also  originated  with  the  state,  but  it  was  nonetheless  real. 
Demand  for  bank  loans  was  not  forthcoming,  because  expansion  in  lines 
other  than  those  which  government  approved  and  largely  financed  was 
impossible.  For  the  same  reason — which  here  stands  out  still  more 
clearly  because  of  the  prohibition  of  all  but  government-approved  issues 
— the  late  and  weak  revival  in  new  industrial  bond  issues  does  not  prove 
anything.  They  were  insignificant  until  1937,  when  they  amounted  to 
258  millions.  Issues  of  shares  increased  steadily  from  the  minimum  in 
1933,  but  amounted  to  only  395  in  1936  (333  in  1937;  the  maximum  was 
1438  in  1927).  But  no  inference  about  absence  or  weakness  of  the  entre- 
preneurial impulse  follows  under  the  circumstances. 

F.  Recovery  and  Recovery  Policy  in  the  United  States  from  1933  to 
1935. — What,  according  to  our  schema,  should  have  been  a  Juglar 
recovery  covers  the  period  from  the  autumn  of  1932  to  the  first  months 
of  1935.2  Separate  treatment  is  necessary,  not  only  because  of  the 
difference  in  phase,  but  because  of  the  presence  of  another  difference 
which  while,  strictly  speaking,  but  one  of  degree  is  so  important  as  to 
amount  to  one  in  kind:  while,  as  we  have  seen,  recovery  policy  was  a 
distinctly  minor  factor  in  1931  and  not  of  decisive  though  of  greater 
importance  in  1932,  it  thenceforth  dominates  the  scene.  This  is  so  obvi- 
ous as  to  raise  the  question  whether  there  is  any  sense  at  all  in  going  on 
speaking  of  cyclical  phases  and  trying  to  date  them,  or  to  relate  actual 

1  The  Reichsbank's  holding  of  bills  other  than  ordinary  treasury  bills  was,  for  the  same 
two  dates,  2,439  and  5,637  millions. 

2  More  precisely,  our  experimental  count  gives  from  the  middle  of  November  1932 
to  March  1935  inclusive,  and  within  this  period  a  Kitchin  recession  to  the  end  of  August 
1933,  a  Kitchin  depression  to  the  middle  of  June  1934,  and  a  Kitchin  revival  covering  the 
remaining  months.     Again,  this  is  not  intended  even  as  a  reference  schema  but  merely  as 
an  illustration.     But  it  should  be  observed  that,  properly  understood,  it  is  not  out  of  step 
with  actual  events. 


984  BUSINESS  CYCLES 

business  situations  to  our  process.  Many  economists  would  not  hesitate 
to  answer  in  the  negative.  Theories  have,  in  fact,  been  offered  which 
are  explicitly  or  implicitly  based  on  the  hypothesis  that  either  from  1914 
or  from  1929 — the  beginning  according  to  some  writers  of  a  series  of 
completely  new  vicissitudes  of  capitalism  that  are  unheard  of  in  the 
sense  that  previous  history  does  not  present  anything  at  all  comparable — 
or,  finally,  from  1933,  a  new  economic  pattern  has  more  or  less  suddenly 
emerged  which  superseded  the  previous  one  for  good  and  calls  for  a  new 
analytic  model  and  fundamentally  new  assumptions  as  to  both  data  and 
mechanisms,  especially  with  respect  to  the  investment  process. 

We  are  not  concerned  with  the  methodology  underlying  those  theories 
and  with  their  intrinsic  merits  or  demerits.1  We  are  face  to  face  not 
with  a  question  of  principle  but  with  a  question  of  fact.  The  only  princi- 
ple involved  is  the  one  which  has  been  stressed  throughout  this  book  and 
which  rests  on  the  certainty  that  the  economic  process  of  capitalist  society 
will  eventually  turn  into  something  fundamentally  different  and  on  the 
ever-present  possibility  that  our  process  be  temporarily  blotted  out  by 
the  action  of  more  powerful  factors,  such  as,  for  instance,  ruled  events  in 
Germany  1914  to  1923.  As  to  the  facts,  we  not  only  know  that  all  the 
essential  features  of  the  postwar  period  up  to  the  world  crisis,  but  also 
those  of  the  world  crisis  itself  answer  perfectly  to  expectation  from  our 
model,  i.e.,  from  past  experience.  Moreover,  we  know  something  else 
that  sweeps  from  our  path  what  otherwise  would,  as  we  know  from  the 
theoretical  discussion  in  Chap.  IV,  be  an  extremely  thorny  problem : 
we  need  not  ask  whether  the  system  "could  have"  recovered  without 
political  action  stimulating  it  out  of  a  state  of  prostration.  For  it  did.2 

1  The  writer's  opinion  about  them  is  contained  in  the  argument  of  this  book  as  a  whole 
and  cannot  be  put  into  a  few  words.     But  it  is  useful  to  remember  that  that  question 
divides  up  into  the  questions  of  logical  correctness  of  those  models  and  of  their  "fit,"  which 
are  entirely  independent  of  each  other.     A  pattern  of  reality,  its  meaning,  and  its  tendencies 
may  be  correctly  seen  by  an  economist  who  yet  blunders  in  constructing  his  theoretic 
model.     And  a  model  may  be  entirely  correct  in  itself  and  yet  fail  to  fit  the  facts  and 
especially  the  tendencies  enshrined  in  contemporaneous  fact.     The  present  writer  feels 
that  if  he  were  at  this  moment  to  attempt  the  task  of  comprehensive  criticism  of  current 
theories,  he  would  have  to  stress  what  seems  to  him  a  large  amount  of  error  of  both  kinds, 
partly  resting  on  unwarranted  generalization  from  recent  experience.     But  he  also  suspects 
that  if  he  were  to  attempt  this  task  10  years  hence,  he  would  have  to  defend  the  kernel  of 
truth  that  in  both  those  respects  these  theories  contain.     Comments  on  some  of  those 
theories  will  be  offered  at  the  end  of  Sec.  G. 

2  The  author,  who  so  often  is  painfully  aware  of  the  fact  that  his  argument  has  to  con- 
tend against  a  powerful  aversion  to  its  real  or  supposed  implications,  thoroughly  enjoys  the 
psychological  vantage  ground  over  which  he  is  traveling  at  this  point.     For  any  aversion 
of  readers  against  accepting  the  writer's  opinion  will  be  much  mitigated  by  the  only  alterna- 
tive open  to  them,  which  to  many  would  be  no  less  distasteful:  whoever  refuses  to  believe 
that  the  recovery  in  1932  occurred  in  the  ordinary  course  of  the  working  of  the  system  will 
have  to  believe  instead,  that  Mr.  Hoover  turned  the  tide.     That  his  administration  would, 


THE  WORLD  CRISIS  AND  AFTER  985 

This  being  so,  affirmative  answer  to  our  question  is  unavoidable: 
there  is  not  only  point  in  going  on  to  relate  the  course  of  events  to  our 
process — in  the  sense  that  we  assume  every  one  of  the  successive  situa- 
tions to  have  been  the  resultant  of  the  working  of  this  process  and  of  the 
effects  of  government  action,  both,  of  course,  not  only  superimposing 
themselves  on  but  also  influencing  each  other — but  there  is  no  choice 
but  to  do  so.  For  it  would  be  contrary  to  all  experience  and  com- 
mon sense,  though  of  course  no  logical  impossibility,  that  a  process, 
which  can  be  strictly  proved  to  have  been  running  its  course  since  at 
least  the  sixteenth  century  and  right  to  the  end  of  1932,  should  have 
come  to  a  stop  suddenly  on  Mar.  4,  1933.  It  should  be  observed,  how- 
ever, how  severely  our  task  restricts  the  scope  of  our  discussion  of  the 
policies  of  the  period.  What  are  their  most  important  aspects  to  many — 
to  those  in  particular,  who  welcome  them  as  the  dawn  of  an  epoch  of  social 
reconstruction — must,  if  we  are  to  focus  our  attention  on  the  mere  effects 
of  those  policies  on  the  process  which  is  the  subject  of  this  book,  be 
excluded  to  the  point  of  exposing  the  writer  to  the  indictment  that  he  is 
completely  lacking  in  social  vision  and  in  the  grasp  of  the  broad  social 
issues  involved.  While  nothing  can  be  done  about  this,1  it  is  hoped  that 
enough  has  been  done  to  protect  the  argument  that  is  to  follow  from  the 
different,  though  cognate,  misunderstanding  which  may  result  from  our 
speaking  of  recovery  policy  as  an  external  factor  acting  on  our  process. 
It  has  been  pointed  out  not  only  that  economic — or  any — policy  grows  out 
of  and  is,  though  not  uniquely,  shaped  by  the  economic  situation  with 
which  it  attempts  to  deal,  but  also  that  in  the  case  before  us  the  short-run 
situation  in  the  spring  of  1933  was  such  as  to  force  any  but  the  strongest 

but  for  the  whim  of  the  political  calendar,  have  come  out  with  flying  colors  and  amidst  a 
universal  clapping  of  hands  is,  for  reasons  of  political  psychology,  very  plausible  in  any  case. 
1  The  very  fact  that  we  refer  to  the  economic  policy  of  the  period  as  recovery  policy 
seems  to  suggest  not  only  narrowness  of  outlook  but  also  the  application  of  an  altogether 
unfair  standard.  Therefore,  it  should  at  least  be  explicitly  stated  that  it  is  fully  recog- 
nized that  those  who  framed  or  defended  those  policies  aimed  at  much  more  than  mere 
recovery  and,  hence,  had  to  strike  a  compromise  between  different  and  in  many  cases 
conflicting  aims.  This  conflict — "reform  versus  recovery" — was  and  is  unavoidable  and 
does  not  in  itself  constitute  a  ground  for  the  charge  of  logical  inconsistency  or  of  error  in  the 
choice  of  means.  But  it  introduced  an  element  of  prejudice  and  insincerity  into  the  discus- 
sion of  the  recovery  problem.  Since  what  a  majority  of  people  really  wanted  was  recovery, 
opponents  of  reforms  developed  a  tendency  to  attribute  to  all  such  measures  effects  prejudi- 
cial to  it,  and  advocates  of  social  reforms,  a  habit  of  wholesale  denial  of  the  presence  of 
that  conflict.  And  it  is  against  theories  framed  for  these  purposes  that  we  have  to  protest. 
Moreover,  in  many  cases  the  element  of  personal  or  group  interest  was  as  obvious  in  the 
arguments  on  the  one  side  as  in  those  on  the  other.  In  still  others,  psychotechnics — 
for  instance,  attacking  an  uncomfortable  piece  of  reasoning  by  impugning  the  motives  or 
sympathies  really  or  supposedly  behind  it — were  more  in  evidence  than  professional 
competence. 


986  BUSINESS  CYCLES 

hands1   irrespectively    of   prevailing   preferences   for   or   aversions   to 
"planning." 

1.  Thus  narrowed  down,  our  task  may  be  further  simplified  by  exclud- 
ing a  number  of  measures  which  cannot  have  had  major  effects  on  succes- 
sive business  situations  or  which  cannot  have  had  them  before  1935. 
Nobody  will,  for  example,  attribute  major  consequences  to  the  Federal 
Economy  Act  of  Mar.  20,  1933,  or  to  the  revision  of  veterans'  claims  (Sec. 
20  of  Independent  Offices  Appropriation  Act  of  June  X6,  1933)  or — 
though  some  stimulating  effect,  extending  beyond  the  industries  directly 
affected,  is  of  course  beyond  doubt  in  this  case — to  the  modification  and 
the  subsequent  repeal  of  prohibition.  Most  of  the  more  than  80  acts 
passed  by  the  Seventy-third  Congress  up  to  June  16,  1933,  may  for  our 
purpose  be  dismissed  as  both  uncontroversial  and  unimportant,  although 
the  sum  total  of  them,  no  doubt,  influenced — mainly  steadied— the  exist- 
ing situation.  The  only  measure  of  this  class  for  which  this  may  be  and 
actually  has  been  called  into  question  is  the  Securities  Act  of  May  27, 
1933.  The  writer  would  have  passed  it  by,  thinking  that  it  was  not 
only  the  kind  of  thing  that  has,  ever  since  South  Sea  Bubbles  days,  often 
been  done  after  abnormally  severe  breakdowns  revealing  reckless  financial 
practice,  but  also  a  sober  and  well-drafted  piece  of  legislation,  from  which 

1  Two  facts  underlie  what  the  reader  will,  according  to  his  pleasure,  call  either  the 
criminal  folly  or  the  profound  wisdom  of  American  recovery  policy.  First,  to  a  greater 
extent  than  is  generally  admitted  in  popular  discussion,  the  measures  taken  in  and  after 
1933  but  continue  and  develop  what  had  been  done  or  begun  before.  Second,  the  measures 
of  social  reform  actually  taken  run  on  familiar  European  lines,  some  inaugurated  by  men  as 
little  radical  as  Bismarck  and  Taaffe  (Count  Edward  Taaffe,  Austrian  prime  minister 
1880  to  1892),  others  outlined  in  the  German  Arbeitsrecht  of  the  twenties.  It  seems  bad 
sociology  to  call  the  United  States  "backward"  in  this  respect,  but  many  of  the  most 
characteristic  New  Deal  measures  would  naturally  be  suggested  by  the  simple  consideration 
that  this  is  no  longer  a  pioneer  country.  Two  more  remarks  on  New  Deal  policy :  a  measure 
does  not  stand  or  fall  with  the  arguments  that  are  used  by  its  promoters  and  cannot,  for 
example,  be  called  foolish  merely  because  those  arguments  are;  and  the  ultimate  aims  a 
measure  is  to  serve  are  not  only  irrelevant  for  our  purpose  but  extremely  difficult,  if  not 
impossible,  to  know.  The  monetary  legislation  of  1933  may  illustrate  this.  It  has — 
even  by  some  of  its  supporters — been  called  inflationary.  We  may  perhaps  feel  fairly 
confident  in  diagnosing  the  aims  of  the  various  "inflationist"  groups — debtors,  speculators, 
silver  men,  exporters,  and  even,  so  it  seems  to  the  writer,  banks  which  while  standing  for 
financial  decorum  were  presumably  not  all  of  them  sorry  to  have  their  credits  unfrozen — 
but  it  is  certainly  beyond  the  writer  to  define  unequivocally  the  subjective  meaning  of  the 
actual  measures  enacted.  Given  the  fact,  if  it  was  the  fact,  that  "inflationist"  interests 
were  at  the  moment  strong  enough  to  override  a  veto  and  to  have  their  way,  the  method 
of  having  "inflationist"  powers  conferred  on  the  president  and  thus  gaining  time  for 
recovery  to  gather  force  and  for  dealing  with  each  of  those  groups  separately  might  con- 
ceivably have  commended  itself  to  a  sound-money  man  of  classical  tradition,  and  no  doubt, 
bears  interpretation  in  an  anti-inflationist  sense,  especially  as  those  groups  were  subse- 
quently, in  fact,  separately  satisfied  in  ways  that  salvaged  considerable  portions  of  that 
classical  tradition.  This  will  be  shown  presently. 


THE  WORLD  CRISIS  AND  AFTER  987 

no  depressive  influence  could  emanate.  But  it  caused  a  flutter,  and  not 
only  in  interested  quarters,  the  main  point  attacked  being  the  construc- 
tion of  the  liabilities  imposed  on  issuing  houses  and  security  dealers. 
These,  however,  do  not  seem  to  amount  to  more  than  responsibility  for 
what  one  "knows  or  ought  to  know"  or  to  enforce  more  than  the  care 
habitually  taken  by  any  decent  firm.  Much  more  plausible  reasons  than 
this  act  are  available  in  order  to  explain  the  stagnation  in  nonpublic 
issues  at  that  time.1 

Recovery  was  substantially  facilitated  by  the  Emergency  Banking 
Act  of  Mar.  9,  1933,  already  mentioned,  which  provided  machinery  for 
the  reopening  of  closed  banks,2  by  the  Banking  Act  of  June  16,  1933, 
which  introduced  a  number  of  important  reforms — the  most  important 
refer  to  strengthening  the  Federal  Reserve  System's  power  over  members, 
particularly  with  a  view  to  regulating  extension  of  credit  for  speculative 
purposes;  to  holding  company  and  security  affiliates;  to  stricter  centraliza- 
tion of  open-market  operations;  to  branch  banking,  and,  for  us  most 
important  of  all,3  to  deposit  "insurance" — and  by  Title  II  of  the  Emer- 
gency Farm  Relief  Act  of  May  12,  1933,  which  dealt  with  the  agricultural 
credit  and  especially  with  the  refinancing  problem.  The  Farm  Mortgage 
Corporation  Act  of  Jan.  31,  1934,  and  the  Home  Owners'  Loan  Acts  of 
June  13,  1933,  and  April  27,  1934,  were  similarly  to  cope  with  another 
thoroughly  frozen  part  of  the  credit  structure  and  thus  also  to  relieve 
the  banking  situation — they  look  quite  conservative  with  the  homage 
they  pay  to  "local  thrift."  We  will  add  further  examples  of  measures 

1  It  has  been  pointed  out  to  the  writer  that  his  interpretation  rests  on  the  hypothesis 
that  the  provisions  of  the  act  are  to  be  administered  reasonably  and  without  vindictiveness 
or  desire  to  victimize  unpopular  interests.     That  is  so.     It  is  also  true  that  ill-founded  suits 
brought  under  the  act  may  impair  the  position  of  a  firm,  even  if  its  action  be  eventually 
found  unexceptionable.     But  whatever  may  be  thought  about  these  and  similar  arguments, 
the  conclusion  that  the  act  was  no  major  impediment  to  recovery  seems  still  to  stand.     For 
an  authoritative  exposition  of  the  opposite  view,  see  C.  J.  Bullock,  The  Securities  Act  of 
1938,  Review  of  Economic  Statistics,  Jan.  15,  1934.     If  long-run  effects  were  pertinent  to 
our  present  purpose,  the  case  for  that  reform  would  be  still  stronger,  however. 

2  It  has  been  mentioned  before  that  this  act  and  the  amendment,  Sec.  136,  of  the  Federal 
Reserve  Act  (Industrial  Advances  Act,  June  19,  1934)  also  provided  for  powers  for  the 
Federal  reserve  banks  to  lend  to  individuals,  especially  industrial  concerns,  but  that  these 
powers  were  not  used  to  a  significant  extent. 

3  That  Act  represents  the  most  systematic  attempt  that  has  been  made  to  deal  with 
what,  from  the  standpoint  of  the  structure  and  technique  of  banking,  always  struck 
observers  as  the  most  obvious  shortcomings  of  the  American  banking  system.     This  larger 
aspect  is  beyond  our  sphere.     But  it  should  be  mentioned  that  the  problem  of  the  small 
and  inefficient  bank  was  attacked  only  indirectly  and  that  the  problem  of  long-term  and 
illiquid  credit  was  not  attacked  at  all,  except  by  making  such  loans  still  easier.     Anything 
suggestive  of  restriction  of  credit,  the  speculative  purpose  alone  excepted,  was  inacceptable 
to  public,  legislators,  and  experts. 


988  BUSINESS  CYCLES 

of  this  type:  the  extension  in  various  directions  (insurance  companies 
especially)  of  the  scope  of  the  Reconstruction  Finance  Corporation's 
assistance  by  the  Acts  of  June  10  and  14,  1933;  the  Emergency  Railroad 
Transportation  Act  providing  facilities  for  consolidations,  rationaliza- 
tions, and  reorganizations;  and  the  United  States  Employment  Service 
Act  of  June  6,  1933.  * 

These  and  similar  measures  did  not  make  recovery.  They  helped  to 
provide  conditions,  institutional  and  other,  rather  than  stimuli,  for  the 
process  of  recovery  to  resume  quickly  after  the  catastrophe  of  the  spring 
of  1933,  by  solving  individual  problems,  removing  impediments  and 
potential  storm  centers,  constructing  safeguards,  allaying  fears,  and,  on 
balance,  improving  the  general  atmosphere — all  of  which  would  other- 
wise have  been  the  source  of  prolonged  difficulties  and  waste.  In 
these  respects  the  combined  effect  of  the  recovery  measures  of  this  type 
must  be  rated  high,  although,  if  they  had  stood  alone,  we  should  have 
no  hesitation  in  speaking  of  a  process  of  recovery  propelled  by  the  forces 
embodied  in  our  model. 

2.  But  we  shall  arrive  at  no  very  different  conclusion  as  regards  the 
two  towering  monuments  of  early  New  Deal  policy,  the  AAA  and  the 
NRA.  All  the  great  and  small  questions  of  principle  that  surround 
them — especially  those  pertaining  to  the  realm  of  welfare  economics2 — 
and  all  the  ultimate  effects  they  had  or  eventually  would  have  had  on  the 
institutional  framework  of  the  economic  process  of  this  country  and  on 
this  economic  process  itself  being  excluded  from  our  discussion,  we  easily 
arrive — discarding  for  the  moment  the  provisions  on  money  inserted 
into  the  act  creating  the  one  and  the  labor  provisions  contained  in  the 
other  act — at  the  result  that,  on  balance,  both  of  them  promoted  recovery 
of  the  usual  type  without  replacing  it  by  a  recovery  that  would  have  to 
be  explained  on  different  principles.  They  certainly  paralyzed,  and 
replaced  by  others,  certain  parts  of  the  ordinary  capitalist  machine  but, 
taking  the  national  organism  as  a  whole,  in  a  way  and  to  an  extent  which 
was  corrective  rather  than  constructive. 

1  The  Farm  Credit  Act  (June  16,  1988)  was  a  measure  of  agricultural  credit  policy  in 
general,  and  has  but  little  direct  bearing  on  the  particular  emergency,  since  the  safeguards 
provided  for  severely  limited  the  granting  of  distress  loans.     The  Tennessee  Valley  Author- 
ity Act  of  May  1933  would  for  us,  but  for  its  possible  bearing  upon  private  power  finance 
and  investment,  simply  come  in  under  the  heading  of  government  expenditure.     The  far- 
reaching  aspects  with  which  public  discussion  has  invested  it  and,  as  it  were,  the  symbolic 
significance  it  has  thereby  acquired  are  not  relevant  to  our  subject.     See  Professor  E.  S. 
Mason  on  the  Power  Aspects  of  the  TVA's  Program,  Quarterly  Journal  of  Economics,  vol. 
50,  p.  377. 

2  Both  cases,  in  fact,  illustrate  well  the  rationale  of  our  distinction  between  welfare  and 
prosperity. 


THE  WOKLD  CRISIS  AND  AFTER  989 

This  is  especially  clear  in  the  case  of  the  agricultural  adjustment 
policy.1  As  we  know  from  previous  discussions,  it  was  to  deal,  unlike 
NRA,  not  with  an  emergency  simply  but  with  an  emergency  in  which  a 
long  development  creating  fundamentally  untenable  conditions  had 
suddenly  come  to  a  head.  Wholesale  liquidation  of  farms,  impinging 
on  an  industrial  unemployment  which  itself  was  unmanageable  for  the 
moment  and  which  had  set  into  motion  a  current  of  remigration  to  the 
land,  would  have  been  the  "automatic"  method  of  restoring  equilibrium. 
The  alternative  to  this — what  we  have  called  "orderly  retreat" — pre- 
cisely implied  temporary  or  even — for  those  who  from  any  of  the  many 
reasons  one  may  have  for  this  desired  to  keep  alive  a  large  and  contented 
farming  population — permanent  preservation  of  disequilibrium  in  the 
agrarian  sector  to  be  unavoidably  financed  by  the  (normal)  surplus  of 
the  industrial  sector.2  To  do  this  would,  ipso  facto,  facilitate  general 
recovery.  It  would  reestablish  something  like  the  previous  processes 
in  the  agrarian  sector  and  the  previous  relations  of  the  agrarian  to  the 
industrial  sector.  Thus  it  would  also  relieve  the  debt  and  banking 
situation  and  thereby  stop  up  a  source  of  actual  and  possible  cumulative 
depressive  effects.  At  the  same  time  it  could  not,  the  relative  financial 
strength  of  the  two  sectors  being  what  it  is,  exert  pressure  on  the  non- 
agricultural  sector  severe  enough  to  open  up  another  such  source  instead. 
This  argument  applies  a  fortiori  in  case  the  means  transferred  were  to  be 
created  ad  hoc  or  taken  from  sources  other  than  that  part  of  nonagrarian 
incomes  which  was  being  currently  spent.  But  it  should  be  observed 
that  it  would  also  apply  if  nonagrarian  consumers'  expenditure  had  been 
reduced  by  an  equal  amount,  i.e.,  if  the  agrarian  Paul  had  really  only 
received  consumers'  dollars  taken  from  the  industrial  Peter.  The  prob- 
lem thus  being  perfectly  clear  and  soluble,  a  very  simple  program  suggests 
itself  of  refinancing  bona  fide  farmers  threatened  by  foreclosure,  of 
nationalizing  the  marketing  of,  and  particularly  the  export  trade  in, 
agricultural  products,  and  of  strictly  planning  production — into  which 
program  measures  of  more  fundamental  adjustment  and  of  further 
rationalization  could  have  been  inserted  at  will.  Most  of  this  being 
out  of  the  question  under  the  Constitution  and  owing  to  the  unprepared- 

1  Since  it  is  impossible  to  do  justice  to  the  subject,  reference  should  be  made  to  the 
treatment  of  the  recovery  aspects  of  AAA  in  a  series  of  publications  of  the  Brookings 
Institution.     See,  in  particular,  J.  S.  Davis,  Wheat  and  the  AAA,  1935;  H.  J.  Richards, 
Cotton  and  the  AAA,  1936;  J.  D.  Black,  The  Dairy  Industry  and  AAA,  1935;  H.  B.  Rowe, 
Tobacco  and  the  AAA,  1935;  E.  G.  Nourse,  Marketing  Agreements  under  the  AAA,  1935; 
and  especially  Nourse,  Davis,  and  Black,  Three  Years  of  the  Agricultural  Adjustment 
Administration,  1937,  particularly  Chap.  XIV. 

2  From  this  standpoint,  much  talk  about  justice,  income  or  price  parities,  in  itself  open 
to  rather  obvious  objections,  acquires  perfectly  sound  meaning. 


990  BUSINESS  CYCLES 

ness  of  public  opinion,  the  well-known  devious  route  was  chosen  (Title 
I  of  the  Emergency  Farm  Relief  Act  of  May  12,  1933),  which  raises  a  long 
string  of  problems  peculiar  to  it,  incidence  and  other  effects  of  processing 
taxes  among  them. 

But  provided  we  agree  that  the  net  effect  was  to  increase  farm  revenue 
considerably,  we  need  not  go  into  those  problems  since  then  the  rest,  i.e., 
the  proposition  that  a  contribution  was  made  to  general  recovery, 
automatically  follows.  In  order  to  take  this  view,  it  is  not  necessary  for 
us  either  to  accept  palpable  exaggerations  of  the  role  that*  the  farmers' 
plight  played  in  the  general  depression,  such  as  a  prominent  authority 
has  been  guilty  of  in  stating  that  60  per  cent  of  all  the  unemployed  "lost 
their  jobs  because  of  the  reductions  in  rural  buying  power";  or  to  share 
the  opinion  of  fervent  advocates  about  the  effects  of  agricultural  recovery 
on  general  recovery,  some  of  whom  went  so  far  as  to  call  the  advantages 
accruing  to  agriculture  a  mere  "incident"  of  a  general  benefit  conferred 
on  the  nation;  or  to  fall  back  upon  doubtful  theories  about  effects  through 
an  incident  decrease  in  savings  or,  somewhat  less  incorrectly,  nonspend- 
ing;  or,  finally,  to  overlook  the  role  of  the  droughts,  of  the  depreciation 
of  the  dollar,  and  of  general  recovery  itself.  Net  results,  which  are  all 
we  have,  are  at  best  difficult  to  interpret,  and  the  panegyrics  of  the 
administration  of  the  agricultural  adjustment  act  on  its  own  activities 
no  doubt  bear  discounting.  If,  for  example,  we  read  in  its  first  report 
that  in  rural  communities  delinquent  taxes  were  being  paid,  debts  owing 
to  banks  discharged,  schools  reopened,  orders  placed  for  clothing,  fur- 
nishings, implements,  automobiles  and  parts,  all  because  of  AAA,  we 
cannot  help  feeling  that  a  sound  case  is  being  spoiled  by  overstatement. 
It  is,  moreover,  not  easy  to  determine  how  much  of  the  results  actually 
attained  is  attributable  to  the  basic  idea,  restriction  of  production — 
"paying  for  not  producing" — and  how  much  to  other  devices  that  were 
not,  or  not  necessarily,  bound  up  with  it,  such  as  marketing  agreements 
and  semimonopolistic  export  practices.  The  energetic  tobacco  program 
supported  by  the  Kerr-Smith  Act  scored  the  most  striking  success,  in 
fact,  considerably  beyond  the  goal  envisaged,1  owing  to  particularly 
favorable  conditions  of  demand.  In  dairying  where,  because  of  a 
relatively  sound  fundamental  situation,  there  was  much  less  need  for 
action — leaders  accepted  what  was  offered  to  them  rather  than  pressed 
for  it,  and  there  was  more  opposition  to  than  support  for  a  produc- 
tion program — restriction  (purchase  of  dairy  cattle)  was  secondary, 
and  market  agreements — in  some  cases  supplemented  by  what  amounted 
to  internal  import  prohibitions — mainly  did,  not  always  successfully, 

1  That  goal  itself  was  not  defined  as  in  the  other  cases.  When  it  turned  out  that  the 
base  period  1910  to  1914  yielded  a  "parity**  which  was  already  attained  at  existing  prices, 
reasons  were  found  for  a  higher  "fair  exchange  value." 


THE  WORLD  CRISIS  AND  AFTER  991 

what  there  was  to  be  done.1  Leaving  aside  the  somewhat  equivocal 
results  of  the  corn-hog  action  and  other  items,  we  will  recall  that  the 
effect  of  the  wheat  program,  whatever  it  might  have  been,  was  largely 
absorbed  by  the  dominant  effects  of  four  successive  bad  harvests — 1933 
to  1936 — so  that,  disregarding  various  supplementary  measures  of  minor 
importance,  the  actual  benefit  to  wheat  farmers  from  AAA  substantially 
reduced — except  perhaps  for  19342 — to  the  benefit  payments  financed 
by  the  processing  tax  which  for  the  three  years  during  which  the  arrange- 
ment was  in  force,  amounted  to  $326  million  and  may  be  likened  to  a 
simple  subsidy. 

The  cotton  program  was  but  to  a  minor  degree  interfered  with  by 
adverse  natural  conditions.  Participation  was  extensive  from  the  outset 
and  became  still  more  so  under  the  pressure  of  the  Bankhead  Act  and  of 
various  privileges  (seed  loans,  loans  of  the  Commodity  Credit  Corpora- 
tion) that  were  confined  to  participants.  This  actually  did  for  cotton 
what  nature  did  for  wheat.  Curtailment  of  acreage  for  1934  and  1935 
was  made  more  effective  by  destruction  of  roughly  £5  per  cent  of  the 
1933  crop.  (Annual)  prices  to  growers  for  1934  were  nearly  double  the 
(annual)  prices  for  1932 — which  were  depressed,  however,  by  the  govern- 
ment holdings  of  over  3  million  bales  resulting  from  the  operations  of  the 

lSee  the  appraisal  of  J.  D.  Black  in  The  Dairy  Industry  and  AAA,  1935. 

2  Weather  and  restriction  of  acreage — more  precisely  voluntary  restriction — were,  of 
course,  both  "external  factors  impinging  on  the  process  of  recovery"  and  acted  in  part 
alternatively  and  in  part  cumulatively.  But  it  is  held  that  even  in  the  latter  case  weather 
was  the  more  important  factor.  This  may  be  and  has  been  questioned  for  1934,  exponents 
of  the  administration  claiming  that  AAA  accounted  in  that  year  for  a  reduction  of  the  crop 
by  between  50  and  60  million  bushels,  or  about  10  per  cent.  If  this  be  so,  then  the  state- 
ment in  the  text  would  have  to  be  modified  for  that  year,  10  per  cent  being  enough  to  make 
a  very  considerable  difference  to  price  and  value.  Professor  Davis  (Wheat  and  the  AAA, 
p.  349)  estimated  the  difference  made  by  AAA  to  acreage  sown  at  only  about  5.4  per  cent. 
But  even  this  would  be  far  from  negligible.  It  should  be  added  that  it  would  be  nothing 
for  the  administration  to  take  pride  in  if  they  had  intensified  the  effects  of  the  drought, 
nor  a  criticism  of  the  AAA  policy  if  they  had  failed  to  do  so,  but  that  on  the  contrary 
the  latter  ought  to  be  recorded  to  the  credit  of  both  the  arrangement  and  its  handling 
by  the  administration.  The  writer  believes  that  (again  neglecting  effects  of  supple- 
mentary measures  about  exports  and  surplus  clearing  in  the  Northwest)  the  drought 
and  possibly,  for  1934,  AAA  restriction,  together  with  general  recovery,  substantially 
account  for  the  behavior  of  wheat  prices,  and  that  Dr.  Davis  (op,  cit.,  p.  366)  overestimates 
the  role  of  monetary  policy.  The  striking  rise  that  occurred  from  May  to  July  1933  was 
certainly  due  to  a  speculative  movement  induced  by  inflationary  talk,  as  well  as  by  antici- 
pation of  a  bad  harvest  and  of  further  government  action  for  the  benefit  of  agriculture. 
But  the  subsequent  failure  of  prices  to  fall  to,  say,  the  1931  average  cannot,  except  to  a 
minor  extent,  be  ascribed  to  monetary  causes,  because  of  the  inadequate  importance  of 
the  connecting  link,  exports.  We  do  not  deny  that  monetary  policy  played  some  role  and 
shall  mention  it  again  below.  But  it  is  going  too  far  to  attribute  to  it  "something  like  one- 
half"  of  the  increase  from  1932  to  1933-1934. 


992  BUSINESS  CYCLES 

Federal  Farm  Board  1929  to  1931,  to  the  point  of  reconquering  foreign 
markets  lost  (partly)  through  the  price-pegging  policy  of  those  years — 
and  revenue  of  growers  rose  from  483  millions  to  (inclusive  of  benefit 
payments)  880  millions  in  1933  and  893  millions  in  1934.  This  was  not 
even  all.  For  without  the  program,  the  crop  of  1933  would  have  been 
one  of  the  largest  on  record.  In  this  case,  of  course,  monetary  policy 
also  counted  for  more  (see  below)  owing  to  the  relative  importance  of  the 
export  interest.1 

Net  costs  and  losses  to  the  Treasury  (including  reduction  of  the  import 
duty  on  sugar)  were  about  900  millions — about  one-third  of  this  being 
due  to  the  invalidation  of  the  processing  tax — which  may  roughly  be 
said  to  constitute  additional  expenditure  within  the  system.  Apart  from 
this,  such  additional  expenditure  was  involved  in  the  running  of  the 
scheme  though  it  was  originally  intended  to  be  self-supporting.  What- 
ever we  may  think  about  technique,  details,  aims  professed,  or  arguments 
used,  the  success  of  the  policy  in  removing  a  major  obstacle  from  the 
road  of  recovery  and  in  reviving  shriveled  tissues  in  the  economic  organ- 
ism is  beyond  reasonable  doubt. 

3.  Title  I  of  the  National  Industrial  Recovery  Act  of  June  16,  1933, 
as  embodied  in  the  codes  of  fair  competition,  introduced  a  type  of  state- 
supervised  industrial  self-government  the  gist  of  which,  stripped  of 
phraseological  mimicry  and  apart  from  provisions  about  labor,  was  legal 
recognition  and  official  encouragement,  amounting  to  compulsion,  of  a 
modified  form  of  the  German  cartel  which,  quite  independently  of  this 
legislation,  tended  to  grow  out  of  the  activities  of  trade  associations. 
Thus  it  is  not  easy  to  understand — except  on  the  supposition  that  the 
mere  word  Planning  and  the  mere  sight  of  a  government  office  are  to  some 
people  irresistibly  attractive — the  enthusiasm  with  which  some  "liberal" 
economists  greeted  a  measure  which  associated  Planning  with  that  very 
restriction  and  price  rigidity  that  are  usually  debited,  as  its  greatest 
blemishes,  to  the  account  of  Big  Business.  But  it  is  all  the  easier  to 
understand  how  it  helped  recovery:  exactly  as  the  German  cartel, 
it  pegged  weak  spots  within  industries,  stopped  spirals  in  many  places, 
mended  disorganized  markets,  especially  in  cases  of  inelastic  demand  and 
of  that  "overproduction"  which  is  incident  to  the  process  of  underselling 

1  The  Cotton  Option  Plan  of  the  first  year  and  the  10-cent  and  12-cent  loan  policy  had 
effects  on  the  international  position  of  the  United  States  cotton  to  some  extent  similar  to 
those  of  the  Farm  Board's  pegging  policy,  which  preceded  it.  On  this  and  other  points 
see  Professor  Black's  article,  The  Outlook  for  American  Cotton,  Review  of  Economic  Sta- 
tistics, Mar.  15,  1985.  The  present  writer,  being  in  any  case  more  pessimistic  than  Profes- 
sor Black  about  the  long-run  possibilities  of  American  cotton  exports,  does  not  rate  very 
highly  the  unfavorable  effects  which  policies  of  the  type  discussed  may,  in  the  long  run, 
have  on  them.  We  are,  be  it  repeated,  merely  speaking  of  immediate  effects  during  that 
recovery. 


THE  WORLD  CRISIS  AND  AFTER  993 

the  obsolete.  There  is  little  doubt  about  its  effectiveness  in  paralyzing, 
in  some  instances,  the  process  of  industrial  transformation  that  was  going 
on:  the  failure  to  see  that  there  was  such  a  process  at  all  or,  in  fact, 
anything  else  but  breakdown  and  deadlock  was  part  of  the  philosophy 
of  the  time.  There  is  as  little  doubt  about  its  effectiveness  in  improving, 
in  other  instances,  situations  in  which  lack  of  organization  really  wrought 
wanton  destruction  as  in  the  cases  of  oil  and  bituminous  coal1  and  others 
to  which  a  less  sweeping  and  spectacular  measure  could  have  confined 
itself. 

Immediate  results  for  the  general  business  situation  were,  however, 
only  the  stronger  because  of  the  range  over  which  this  policy  was  indis- 
criminately applied — although  very  unequally  enforced — and  so  were  its 
purely  psychological  effects,  which  in  a  situation  of  that  kind  we  have  a 
right  to  consider  an  important  factor — even  Blue  Eagles  do  count  for 
something  when,  objective  conditions  for  revival  being  given,  it  is  broken 
morale  that  is  the  matter.  Invalidation  by  the  Supreme  Court  (June 
1935)  came  when  the  end  had  been  achieved  and  was  for  the  administra- 
tion a  blessing  in  disguise.  But  aftereffects  were  not  entirely  eliminated 
thereby.  Business  had  learned  a  lesson.  The  "chiseler"  continued  to 
be  frowned  upon.  And  we  shall  have  to  bear  in  mind  that  there  is  here 
a  possible,  if  only  partial,  explanation  of  the  fact  that  output  figures 
failed  to  come  up  to  expectation  in  the  subsequent  prosperity  phase, 
which  precisely  for  this  reason  most  economists  preferred  to  call  imperfect 
recovery.2 

1  Section  9  of  the  Act  dealt  specifically  with  oil  restriction.     If  it  did  not  do  so  in  the 
manner  of  the  Kaligesetz,  this  was  presumably  only  due  to  lack  of  powers  and  the  necessity 
of  avoiding  constitutional  pitfalls.     In  the  bituminous  coal  industry  technological  advance, 
increasingly  economical  use  of  coal,  and  increasing  use  of  other  fuels  had  created,  ever  since 
the  war,  conditions  which  afford  a  class-room  example  both  for  the  necessity  and  the  diffi- 
culty of  regulation.     In  1933  and  1934  price  control  was  tried,  but  it  lapsed  with  the 
Supreme  Court's  invalidating  the  Bituminous  Coal  Conservation  Act  (1935). 

2  The  Report  of  the  President's  Committee  of  Industrial  Analysis  formed  to  prepare  a 
"review  of  the  effects  of  the  administration  of  Title  I  of  the  National  Industrial  Recovery 
Act"  after  the  latter  had  been  declared  unconstitutional,  is  not  in  itself  very  interesting. 
But  the  five  studies  of  the  committee's  staff  (unpublished  but  open  to  consultation)  and 
many  official  and  nonofficial  studies — especially  those  that  emanated  from  the  Research 
and  Planning  Division — making  use  of  materials  collected  under  the  administration  of  the 
Act,  are  not  only  an  important  source  of  historical  material  but  also  a  storehouse  of  prob- 
lems and  patterns  for  the  analytical  economist  quite  inadequately  exploited  so  far.     The 
general  literature  of  the  subject  of  NRA  suffers  both  from  insufficient  foundation  in  the 
details  of  actual  fact  and  from  the  preconceptions  of  authors,  which  in  the  majority  of 
cases  are  obvious  from  the  start  and  even  affect  statements  of  fact.     A  well-chosen  sample 
of  the  more  significant  publications  has  been  ably  discussed  by  Professor  Rogin  in  The  New 
Deal,  Quarterly  Journal  of  Economics,  February  1935.     But  such  study  as  the  present 
writer  has  been  able  to  make  has  convinced  him  that  for  the  very  limited  purpose  in  hand 
the  view  expressed  in  the  above  paragraphs  would  be  substantially  accepted  by  a  majority 


994  BUSINESS  CYCLES 

The  quid  pro  quo  which  the  Act  and  the  Codes  offered  to  possible 
opposition  from  the  ranks  of  labor  and  to  possible  criticism  of  what  might 
easily  have  been  called  antisocial  tendencies,  were  the  labor  provisions. 
Many  an  objector  was  reconciled  by  the  clauses  on  child  labor,  hours,  and 
minimum  wages  which  all  codes  (including  the  blanket  code)  contained, 
or  gratified  by  the  protection  extended  in  Sec.  7a  to  collective  bargaining, 
to  organizing  activity,  and  so  on,  which  went  far  beyond  the  Norris- 
LaGuardia  Anti-Injunction  Act  of  1932.  The  larger  aspects  of  the 
progress  thereby  achieved  in  social  legislation  being  once  more  beyond 
our  scope — of  a  progress  that  many  of  us  will  consider  to  have  been 
overdue — our  only  question  concerns  the  effect  on  recovery  of  general 
labor  policy  and  of  wage  policy  in  particular.  As  to  the  first,  the  present 
writer  is  confident  that  no  inhibiting  effect  can  be  proved  for  the  period 
under  consideration.1  As  to  the  second,  it  follows  from  previous  argu- 
ment that,  under  the  conditions  of  this  country  and  of  the  prevailing 
cyclical  phrase,  the  persistent  official  efforts  to  raise  the  whole  structure 
of  wage  rates  must  on  balance  have  had  an  adverse  effect  both  on  the 
expansion  of  output  and  on  employment  per  unit  of  output.  While  this 
effect  was  probably  small  during  the  first  great  upward  rush  in  1933,  the 
further  development  of  output,  pay  rolls,  and  especially  employment, 
which  can  hardly  be  said  to  come  up  to  expectation,  substantiates  the 
presence  of  this  brake  (see  below,  sub  6).  The  reader  will  realize  that 
this  is  perfectly  compatible  with  an  opposite  result  in  many  individual 
cases;  with  such  truth  as  there  is  in  antisaving  arguments;  and  with 
recognition  of  the  facts  that  the  rise  in  price  level  partly  absorbed  and 
government  spending  partly  counteracted2  that  effect.  Again,  and 

of  the  economists  who  worked  in  that  field,  barring  perhaps  the  analogy  with  the  German 
cartel  and  the  last  sentence.  The  reader  is  referred  to  Professor  E.  S.  Mason's  admirable 
contribution  on  Controlling  Industry  in  The  Economics  of  the  Recovery  Program  (D.  V. 
Brown  and  others,  1984),  and  to  L.  S.  Lyon  and  others,  The  National  Recovery  Administra- 
tion: an  Analysis  and  Appraisal,  1935. 

1  In  other  words,  the  mere  recognition  of  the  right  of  collective  bargaining  and  the 
elimination  of  the  yellow-dog  contract  cannot  in  themselves  have  worked  against  recovery. 
It  does  not  follow  that  an  organized  drive,  facilitated  by  the  newly  acquired  opportunities 
for  action,  or  further  legislation  on  the  same  lines  might  not  have  done  so.     But  no  such 
drive  or  legislation  occurred  during  the  two  years  during  which  the  NRA  regulations  of 
wages  and  hours,  paralleled  by  regulations  by  other  federal  agencies  and  state  govern- 
ments, were  in  force.     The  question  will  hence  have  to  be  touched  upon  again  in  the  next 
section. 

2  Since  government  spending  was  in  part  a  function  of  the  amount  of  unemployment, 
it  can  even  be  said  that  any  unemployment  created  (relatively  to  what  unemployment 
there  would  otherwise  have  been)  by  that  wage  policy  may  possibly  have  induced  a  net 
increase  of  total  expenditure  in  the  system.     It  will  be  seen,  however,  that  this  does  not 
invalidate  our  argument.     The  Research  and  Planning  Division  of  the  NRA  is  responsible 
for  a  valuable  source  book  (mim.,  January  1935)  on  Hours,  Wages  and  Employment  under 
the  Codes. 


THE  WOKLD  CBISIS  AND  AFTEH  995 

notwithstanding  the  rise  in  price  level,  labor  was  made  expensive  rela- 
tively to  real  capital.  Coupled  with  a  cheap  money  policy,  a  high-wage- 
rate  policy  was,  under  the  circumstances  of  phase  and  country,  the  very 
recipe  for  the  production  of  a  maximum  of  unemployment.  It  should  be 
observed,  however,  that  while  dampening  recovery,  this  would  not 
necessarily  affect  the  general  character  and  the  duration  of  the  phase. 

Let  us  pause  for  a  moment  in  order  to  take  stock.  So  far  we  have 
before  us  the  following  elements  of  the  situation  in  1933.  There  was 
first  incipient  recovery  dating  from  about  the  middle  of  1932.  This 
incipient  recovery,  second,  had  been  interrupted  by  the  banking 
catastrophe  in  the  spring  of  1933  and  was  weighed  down,  independently 
of  that,  by  the  conditions  in  the  agrarian  sector,  by  the  results  and 
remains  of  a  preceding  state  of  overindebtedness  in  general,  and  by 
depressive  factors  special  to  certain  individual  industries.  Third,  on  this 
pattern  impinged  the  series  of  measures  mentioned,  all  of  which,  with  the 
exception  of  the  high-wage-rate  policy  inaugurated — though  this  was  of 
small  immediate  importance  in  1933- — were  on  balance  remedial  in  effect, 
i.e.,  not  only  devised  in  order  to  remove  those  millstones  but  actually 
effective  in  achieving  this.  Nothing  more  than  these  three  groups  of 
facts  is  necessary  for  us  to  expect  a  strong  and  even  violent  rebound  of  the 
system,1  more  than  compensating  for  the  subnormal  revival  during  the 
preceding  six  months,  to  be  interrupted,  however,  by  the  setting  in  of  a 
Kitchin  depression  late  in  the  year. 

This  may  be  expressed  by  means  of  terms  which  have  been  so  uncriti- 
cally used  as  to  elicit,  when  used  now,  little  else  but  contempt:  natural 
and  sound  recovery.  To  the  former  term  we  assign  the  meaning  of 
recovery  coming  about  in  the  course  of  the  cyclical  process  by  virtue  of 
its  mechanism.  The  latter  term  we  define  as  a  recovery  that  is  brought 
about  by  factors  which  do  not  carry  an  inherent  tendency  to  reproduce  the 
same  difficulties  as,  or  to  produce  other  difficulties  in  place  of,  those  that 
they  have  been  instrumental  in  overcoming:  relapse  is  the  most  obvious 
instance  of  the  first  type,  and  undoing  of  such  work  of  readjustment  as 
may  result  from  the  processes  of  depression  is  an  instance  of  the  second. 
It  will  be  seen  that  natural  and  sound  recovery  are  not  made  synonymous. 
Whether  "natural"  recovery  is  always  "sound"  depends  on  whether  we 
exclude  all  those  "understandable  nonessentials "  which  may  easily  land 
the  system  in  an  untenable  situation.  But  inasmuch  as  depression  itself 

1  That  statement,  forming  as  it  does  an  important  link  in  our  diagnostic  argument, 
should  perhaps  be  amplified  and  buttressed.  But  the  writer  does  not  see  how  this  could 
be  done  without  intolerable  repetition.  It  should,  however,  be  obvious  that  stopping 
spirals  and  reversing  "psychology"  would,  in  a  situation  fundamentally  ready  for  improve- 
ment and  with  levels  only  un9erstandable  from  spirals  and  psychology,  produce  a  strong 
spurt. 


996  BUSINESS  CYCLES 

is  a  pathological  process,  sound  recovery  need  not  be  natural.  In  our 
case  the  recovery  we  now  envisage  would — whatever  might  have  been 
the  ultimate  effects  of  NRA — have  been  substantially  sound.  And  a 
natural  recovery  was  at  the  bottom  of  it.  But  the  midwife  role  of  public 
authority  was  so  important  that  it  would  not  do  to  draw  laissez-faire 
conclusions. 

It  is  not  implied  that  Congress  or  the  administration  "should"  or 
"could"  have  stopped  at  the  measures  so  far  mentioned.  The  prevail- 
ing unemployment  would  in  itself  suffice  to  refute  any  such  implication. 
But  it  is  implied  as  a  matter  of  diagnosis  that  to  a  greater  extent  than  is 
commonly  realized  the  recovery  of  1933  can  be  accounted  for  irrespec- 
tively of  the  monetary  and  spending  policies  inaugurated  at  that  time. 
For  this  proposition  there  is,  it  is  true,  no  such  proof  as  the  constellation 
of  facts  enabled  us  to  give  for  the  cognate  proposition  that  recovery 
started  "of  itself"  and  not  simply  in  response  to  recovery  policy.  But 
there  is  a  prima-facie  case  which  is  much  strengthened  by  the  fact  that 
during  the  critical  second  and  third  quarters  of  1933  the  new  spending 
program,  being  still  in  the  incubating  stage,  cannot  have  had  major,  at 
least  mechanical,  effects1  on  the  economic  process.  That  substantially 
the  same  holds  true  of  the  monetary  policy  will  be  evident  from  the 
cursory  discussion2  to  which  we  now  turn. 

4.  When  banks  reopened  after  the  "holiday"  in  March  1933,  they 
were  estopped  from  paying  out  gold  except  by  special  Treasury  per- 
mission— redemption  of  notes  in  gold  was  discontinued  on  Mar.  4 — but 
only  an  insignificant  fall  in  the  international  value  of  the  dollar  occurred. 
On  the  afternoon  of  Apr.  19,  i.e.,  immediately  after  the  declaration  of  the 
(partial)  embargo  on  gold,  dollars  were  internationally  traded  in  at  a 
discount  of  about  9  per  cent  and  during  the  following  eight  days  at  a  dis- 
count which  fluctuated  between  8  and  12  per  cent.  There  was  no  panic 
such  as  had  occurred  in  the  English  case.  This  is  highly  significant.  It 
shows,  on  the  one  hand,  that  international  speculation  did  not  believe  in 
a  substantial  depreciation  of  the  dollar — in  spite  of  the  fact  that  a  bill 

1  According  to  Mr.  Currie's  and  Mr.  Viilard's  figures,  published  by  Mr.  Arthur  D. 
Gayer  (op.  cit.),  net  federal  income-generating  expenditure  for  the  year  was  1,856  millions, 
only  210  millions  more  than  in  1982.     If  the  effect  was  more  than  proportionately  greater 
than  this,  it  can  only  have  been  due  to  the  difference  in  cyclical  phase. 

2  Strictly  confined  to  the  limits  set  by  our  purpose,  that  discussion  not  only  neglects 
details  and  technical  aspects — such  as  the  sequestration  of  gold,  the  abrogation  of  the  gold 
clause,  the  stabilization  fund,  later  on  the  tripartite  agreement,  and  so  on — but  also  majoi 
problems  which  for  that  purpose  are  of  minor  importance.     The  reader  is  referred  to  the 
literature  on  the  subject,  but  Professor  Harris's  Exchange  Depreciation,  1936,  and  also 
his  article  on  the  American  Experience,  Quarterly  Journal  of  Economics,  August  1934,  should 
be  specifically  mentioned.     We  neglect,  in  particular,  the  silver  policy  although,  owing  to 
the  nonnegligible  importance  of  the  relations  with  silver  countries,  some  effects  on  cyclical 
situations  have  doubtless  been  exerted  by  it. 


THE  WORLD  CRISIS  AND  AFTER  997 

was  introduced  on  Apr.  20  which  contained  the  main  provisions  that  were 
later  embodied  in  the  Thomas  amendment — and  on  the  other  hand,  that 
in  the  absence  of  such  speculative  anticipation  there  was  no  reason 
why  the  dollar,  if  left  to  itself,  should  fall  at  all.  There  was,  thus,  a 
complete  absence  of  parallelism  between  the  fate  of  the  dollar  and  the 
fate  of  the  English  pound.  The  dollar  was  not  under  economic  pressure 
either  in  a  short-run  or  in  a  long-run  sense,  either  from  abroad  or  at 
home.  We  recall  in  particular  that  it  had  weathered  the  preceding 
storm,  the  third  through  which  it  had  had  to  sail  during  the  depression: 
the  reserve  system  had  the  situation  well  in  hand,  and  in  less  than  three 
weeks  after  the  bank  holiday  over  half  of  the  amount  was  repaid  which 
had  previously  been  borrowed  from  reserve  banks  in  order  to  meet  loss  of 
gold  and  withdrawals  of  notes  from  all  centers  and  of  bankers'  balances 
from  New  York.  And  almost  all  the  losses  of  reserves  suffered  by  reserve 
banks  were  made  good. 

What  pressure  there  was,  was  entirely  political,  and  it  was  not  until 
the  world  became  convinced  of  the  imminence  of  "inflation"  that  the 
dollar  really  gave  way.  This  conviction  was  the  result  of  the  passing  of 
Title  III  of  the  Agricultural  Adjustment  Act  of  May  12,  which  was  even 
officially  referred  to  as  Emergency  Relief  and  Inflation  Act.  We  are  not 
going  to  waste  space  on  an  otiose  discussion  about  whether  the  course 
actually  followed  during  the  period  deserves  to  be  called  inflationary  or 
not,  and  if  so,  in  what  sense.  Although  there  cannot  be  any  doubt  that 
the  interests,  the  exponents  of  which  swapped  votes  in  order  to  insure  the 
passage  of  that  act,  aimed  at  inflation  in  every  conceivable  sense  of  the 
term,  the  act  itself  was  the  result  of  a  compromise  that  yielded  less  ground 
to  them  than  it  seemed  to  yield  and  than,  from  a  wide  variety  of  motives, 
almost  everyone  pretended  to  believe.  Its  provisions,  besides  being 
only  enabling  and  not  mandatory,  offered  plenty  of  opportunity,  sub- 
sequently extended,1  to  defeat  any  kind  of  inflation  at  will,  and  effective 
use  was  made  of  it  up  to  1937.  It  was  clear  enough,  however,  that  at 

1  Section  46  of  the  act,  amending  Sec.  19  of  the  Federal  Reserve  Act,  gave  power  to 
the  Federal  Reserve  Board  to  change  the  amount  of  reserve  balances  which  members  are 
required  to  keep  against  deposits.  The  Banking  Act  of  1938  prohibits  members  to  act  as 
agents  for  corporations  and  individuals  in  making  loans  on  securities  (loans  on  account  of 
others).  The  Securities  Exchange  Act  of  June  6, 1934,  gives  the  Board  of  Governors  power 
to  prescribe  margins  for  brokers'  loans  to  their  customers  and  for  security  loans  by  banks. 
The  Gold  Reserve  Act  of  Jan.  30,  1934,  has  been  so  constructed  as  partly  to  eliminate  the 
"inflationary"  effect  of  devaluation.  The  Treasury's  braking  power,  already  great,  has 
then  been  strengthened  by  the  Social  Security  legislation  which  gives  it  control  over  vast 
funds  that  may  be  used  for  the  purpose.  These  are  only  some  of  the  brakes  built  into  the 
machinery.  To  be  sure,  there  are  limits  to  their  effects,  and  most  of  them  must  be  put  into 
operation  by  visible  and  unpopular  action.  They  therefore  do  not  do  away  with  the 
possibility  of  inflation  in  every  sense  of  the  term.  But  it  should  be  obvious  how  risky 
numerical  predictions  about  the  future  behavior  of  the  price  level  must  be. 


998  BUSINESS  CYCLES 

least  devaluation  would  be  unavoidable.  Even  so,  the  dollar  manifested 
its  natural  strength  by  the  hesitancy  with  which  it  fell.  Therefore, 
when  in  the  autumn  recovery  dimmed  and  NRA  and  AAA  enthusiasms 
cooled,  when,  moreover,  the  announcement  by  the  Reconstruction 
Finance  Corporation  of  its  willingness  to  lend  to  banks  up  to  a  billion  for 
the  purpose  of  relending  failed,  as  it  naturally  would,  to  produce  results, 
the  administration  encountered  the  consequent  inflationist  onslaught  by 
resorting  in  October  to  a  method  which  would  bring  $own  the  dollar 
without  "inflation,"  viz.9  the  gold-buying  policy,1  and  by  speedily 
investing  the  political  capital  thus  gained  in  putting  a  stop,  for  the 
time  being  at  least,  to  this  type  of  "experiment"  by  the  Gold  Reserve 
(Devaluation)  Act.2  And  still  the  dollar  resisted:  a  torrent  of  gold 
turned  toward  this  country.  We  will  but  glance  at  the  movements  in  its 
gold  stock  during  the  February  following  upon  the  presidential  proclama- 
tion of  Jan.  31,  which  raised  the  price  of  the  fine  ounce  from  $20.67  to  $35 
and  the  value  of  the  monetary  gold  stock — including  gold  previously 
acquired  by  the  Reconstruction  Finance  Corporation  and  the  Treasury, 
but  excluding  coin  still  reported  as  in  circulation — to  7.03  billions.  No 
less  than  381  millions — a  record — were  imported  during  the  month  (213 
of  which  from  England)  chiefly  in  response  to  the  new  price,  banks  at 
home  and  abroad  taking  advantage  of  the  undervaluation  of  the  dollar. 
There  was  also  some  release  (8.6  millions)  from  earmark. 

In  order  to  appraise  the  effects  of  this  policy  on  the  economic  process, 
it  is  first  of  all  necessary  to  realize  how  much  or  how  little  it  had  to  do  with 
easy  money.  There  is  no  doubt,  of  course,  that  the  influx  of  gold,  which 

1  Inflation  of  the  German  type  would,  of  course,  automatically  send  down  the  dollar. 
It  was  to  association  with  such  inflation  (though  not  necessarily  with  exactly  that  dose  of 
it)  that  part  of  the  support  of  the  policy  of  depreciation  and  devaluation  was  due.     What 
this  group  of  "inflationists"  overlooked  was  the  possibility  of  having  the  token  without  the 
substance.     Although  facilitating  future  inflation,  reduction  of  the  gold  value  of  the  dollar 
by  the  gold-buying  policy — direct  action  on  exchanges  by  offering  unlimited  amounts  of 
dollar  balances  would  have  been  a  still  more  effective  method — was  really  (whatever  the 
intention)  a  means  of  avoiding  it.     Its  tactical  virtue  consisted  in  the  fact  that,  in  spite  of 
being  a  device  to  avoid  inflation,  it  would  satisfy  certain  inflationary  interests,  e.g.,  export- 
ers and  speculators,  who  were  particularly  vocal  and  supplied  the  motive  power  of  much 
of  the  inflationary  propaganda.     Hence,  it  would  break  the  inflationary  phalanx. 

2  It  will  be  observed  that  our  interpretation  runs  in  terms  of  tactics  and  interests  rather 
than  in  terms  of  "theories"  and  "experiments,"  and  severely  disregards  phraseology. 
This,  so  it  seems  to  the  writer,  is  as  it  should  be  if  grasp  of  situations  and  not  phraseology 
be  our  aim.     Nothing  is  implied  about  actual  motivation.     But  that  interpretation  has 
until  1937  been  quite  consistently,  borne  out  by  facts.     Although  inflationary  phrases, 
"theories,"  and  ends  were  freely  made  to  serve  tactical  purposes  later  on,  in  each  case 
pressure  on  the  administration  was  more  obvious  than  pressure  by  the  administration. 
The  incessant,  though  mistaken,  appeals  to  banks  to  lend  freely  acquire  their  significance 
in  this  light.     For  utilization  of  existing  facilities  would  draw  the  ground  from  under  the 
feet  of  advocates  of  "currency  inflation." 


THE  WORLD  CRISIS  AND  AFTER  999 

already  in  February  1934  carried  excess  reserves  of  all  member  banks  to 
the  new  peak  of  above  1  billion,  was  then  and  later  the  chief  factor 
responsible  for  what,  at  least  from  the  spring  of  1935  on,  will  strike  us  as 
abnormally  low  rates.  With  government  expenditure  what  it  was  and 
reviving  business,  for  example  prime  bankers'  acceptances  (90  days) 
could  hardly  have  reached  %  of  1  per  cent  by  November  1934  and  stayed 
there,  nor  prime  commercial  paper  %  to  1  per  cent  by  June  and  fallen  to 
%  after  that.  But  if  such  lows  were  beyond  expectation  from  our  model, 
prevalence  of  very  low  rates  was  not,  and  it  is  not  very  obvious  that,  if, 
for  instance,  New  York  City  customers*  rates  steadily  fell  from  their 
modest  "panic  peak"  of  4.88  for  March  1933,  to  2.64  for  March  1935, 
this  was  substantially  more  than  we  should  have  expected  to  find  without 
the  gold  movements  induced  by  devaluation.  In  other  words,  the  latter 
did  not  create  the  conditions  of  monetary  ease.  The  commercial  paper 
rate  was  down  to  2  per  cent  by  the  autumn  of  1931  and,  in  spite  of  the 
subsequent  rise,  to  less  than  that  by  the  autumn  of  1932,  and  there  is 
nothing  in  the  processes  of  incipient  recovery  to  enforce  an  upturn.  As 
far  as  this  goes,  devaluation  did  not  lift  any  weight  from  the  economic 
process,  as  is  indeed  obvious  from  the  behavior  above  described  of  the 
dollar  in  1933.  Moreover,  such  pressure  on  rates  as  there  was  during 
that  year  was  due  not  only  to  the  depreciation  but  also  to  the  open- 
market  purchases  of  the  Federal  reserve  banks,  which  in  response  to  the 
greenback  threat  contained  in  the  Inflation  Act  acquired  570  millions 
of  governments  from  May  to  November  and  then  stopped  because  this 
only  served  to  swell  excess  reserves.  But  that  step  would  have  been 
possible  without  going  off  gold.  Its  ineffectiveness  finally — a  last  veri- 
fication of  our  views  on  the  subject — goes  far  toward  establishing  the 
proposition  that,  whatever  influence  on  rates  and  credit  facilities  was 
exerted  by  whatever  external  factor,  the  influence  of  these  on  the  economic 
process  was  next  to  nothing. 

Devaluation  must,  in  the  second  place,  be  considered  in  relation  to  the 
policy  of  public  expenditure.  It  has,  in  fact,  been  held  that  the  meaning 
of  the  former  primarily  consists  in  its  implementing  the  latter,  which  was 
what  really  produced  results.  There  is,  of  course,  some  truth  in  this 
view,  which  is  at  any  rate  much  superior  to  the  naive  belief  that  redefining 
the  gold  content  of  the  dollar  would  per  se  change  the  price  level  in  the 
same  proportion — a  curious  survival  from  the  days  of  the  commodity 
theory  of  money.  For  although  increase  in  price  level  neither  is,  as  a 
matter  of  principle,  nor  has  been,  as  a  matter  of  fact  in  this  instance,  the 
main  effect  of  antidepression  public  expenditure,  it  is  true  that  devalua- 
tion can,  with  the  qualification  to  be  mentioned  presently,  only  act  on 
prices  if  and  when  it  either  induces  or  facilitates  increase  in  expenditure. 
Therefore,  if  public  expenditure  that  would  not  otherwise  have  been 


1000  BUSINESS  CYCLES 

technically  possible  had  been  made  so  by  devaluation,  we  should  have 
to  list  the  latter  among  the  major  factors  influencing  the  recovery  process. 
But  public  expenditure  perfectly  adequate  to  produce  the  results  that 
actually  were  produced  would  have  been  possible  with  the  dollar  at  the 
old  gold  par.  Devaluation  may  have  facilitated  it  by  removing  all  con- 
cern about  monetary  limitations,  but  this  is  all. 

There  remain,  in  the  third  place,  direct  effects.  We  will  mention 
two.  The  year  1933  was  one  of  monetary  disorder  ancj  of  widespread 
apprehensions  about  impending  inflation.  An  impulse  was  thereby 
given  to  speculation  in  securities  and  commodities,  which  was  very  obvi- 
ous on  stock  and  produce  exchanges.  Stock  prices  (and  prices  of  second- 
class  bonds),  in  particular,  reacted  visibly  and  until  September  con- 
sistently to  every  drop  of  the  international  value  of  the  dollar.  The  only 
question  is  how  far  this  effect  extended  beyond  speculation  in  a  wide 
sense  of  the  term.  We  have  seen  that  the  response  of  productive  business 
to  monetary  policy  is,  to  say  the  least,  equivocal.  The  notable  instances 
of  1896  and  1878  should  suffice  to  show  that  anticipation  of  monetary 
expansion  is  not  necessarily  a  propelling,  and  anticipation  of  "sound" 
money  not  necessarily  a  depressing  influence.  Nor  is  this  at  all  astonish- 
ing. Hence,  although  the  anticipations  induced  by  going  off  gold 
certainly  contributed  to  the  hectic  rise  of  the  wholesale  price  index 
— which,  as  has  been  emphasized  by  Mr.  Snyder,  reflects  speculative 
movements  so  disconcertingly  well — which  occurred  from  March  to 
July  1933,  and  although  this  spurt  in  turn  no  doubt  had  some  effect  on 
productive  operations,  we  shall  not  weight  that  component  very  highly, 
especially  in  view  of  the  fact  that,  as  pointed  out  before,  strong  revival 
would  in  any  case  be  understandable  without  it. 

The  other  immediate  effect  of  depreciation  and  devaluation  is  on 
foreign  trade.  The  changes  during  our  period  in  raw  figures  of  either 
values  or  quantities — which  are  not  impressive — do  not  in  themselves 
prove  much.  As  to  export  of  United  States  merchandise,  they  are  com- 
patible with  the  opinion  that  the  motorcar  and  machinery  industries 
benefited  somewhat.  Since,  however,  industrial  exports  are  in  any  case 
not  important  enough  to  matter  greatly,  we  need  not  enter  into  the  nice 
questions  surrounding  any  effort  to  isolate  the  influence  of  devaluation. 
The  official  index  of  quantity  of  agricultural  products  exported  continued 
its  downward  course  throughout  the  period,  the  annual  figure  for  1935 
being  a  little  less  than  58  per  cent  of  that  for  1932.  Some  benefit  to 
wheat  and  especially  to  cotton  farming  is,  nevertheless,  beyond  doubt. 
In  the  latter  case  it  also  served  to  counteract  the  effect  of  the  depreciation 
of  the  rupee,  the  Egyptian  pound,  and  the  milreis.  However,  this 
instance  only  strengthens  the  case  for  the  broad  proposition  that,  with 
qualifications  which  need  not  be  repeated,  public  spending  was  the  only 


THE  WORLD  CRISIS  AND  AFTER  1001 

positively  propelling  measure  acting  on  our  process — as  distinguished 
from  the  measures  previously  discussed,  which  mainly  removed  obstacles 
and  thus  (again,  mainly1)  helped  in  what  may  be  termed  a  negative  way. 

5.  Prima-faeie  federal  income-generating  expenditure  actually  was 
from  about  December  1933  to  about  the  middle  of  1937  the  dominant 
factor  in  the  increase  of  net  national  income  (current  dollars2). 

We  will  not  go  into  the  methods  and  agencies  by  which  the  spending 
program  was  put  into  effect  and  which,  along  with  the  immediate  objects 
of  the  spending  activity,  were  in  an  incessant  process  of  change  from  the 
time  that  foundations  were  laid  by  the  Unemployment  Relief  Act  of 
Mar.  31,  1933  (forestation,  prevention  of  soil  erosion,  plant  disease,  and 
so  on),  the  Emergency  Relief  Act  of  May  12,  1933  (revision  of  powers  of 
the  Reconstruction  Finance  Corporation,  creation  of  the  Federal  Relief 
Administration,  grants  to  states  for  the  purpose  of  direct  unemployment 
relief)  and  titles  II  and  III  of  the  National  Recovery  Act  of  June  16 
(Public  Works  Program,  grants  to  states  for  highway  construction, 
appropriation  of  3.3  billions).  Methods  and  objects  are  not  indifferent 
either  for  the  recovery  or  any  other  aspect.  Even  for  the  immediate 
effect  on  the  economic  process  it  is  by  no  means  indifferent  whether  a 
given  sum  be  spent  on  direct  relief  or  on  purchase,  from  stock,  of  materials 
the  proceeds  of  which  are  used  by  recipients  for  repaying  debts :  there  is, 
according  to  the  way  of  spending,  a  continuous  variety  of  effects  which 
range  from  increasing  system  expenditure  by  several  times  the  amount 
spent  down  to  not  much  more  than  refinancing.3  Nevertheless,  we  will 

1  Some,  especially  AAA,  also  added  to  total  expenditure  in  the  system  and  as  far  as 
they  did,  now  come  in  again  by  virtue  of  this  title.     In  a  small  way,  of  course,  all  of  them 
did  so,  owing  to  the  administrative  expenditure  involved. 

2  See  Professor  Kuznets'  National  Income,  p.  8,  Table  1,  Col.  3.     The  figures  for  net 
federal  income-increasing  expenditure  are  again  Mr.  Gayer's  (op.  cit.).     It  should  be 
added  that  Mr.  Currie's  method,  on  which  they  are  based,  includes  also  receipts  from 
estate  taxes,  on  the  ground  that  under  the  circumstances  estates,  even  when  able  to  pay 
those  taxes  from  owned  cash  and  "near-cash,"  would  do  so  from  idle  funds,  which  were 
thereby  put  into  circulation.     It  will  be  seen  that  income-increasing  expenditure  greatly 
differs  from  the  deficit  as  officially  figured.     All  expenditure  incident  to  refinancing  trans- 
actions is  excluded,  while  income-generating  expenditure  of  nominally  independent  agencies 
(trust  funds  and  so  on)  is  included.     Correction  for  variations  in  the  income-generating 
expenditure  of  states  and  local  bodies  would  not  make  a  significant  difference. 

3  Another  point  may  be  mentioned  here.     There  is  a  type  of  public  works  which 
amounts  to  little  more  than  direct  relief.     But  on  the  other  end  of  the  scale  there  is  a 
type  which,  even  if  undertaken  from  the  relief  motive,  is  nothing  but  a  businesslike  reaction 
of  public  bodies  to  the  prices  of  factors  and  the  rates  of  interest  which  obtain  in  a  period  of 
depression  or  incipient  recovery.     Strictly,  expenditure  of  this  character  should  be  excluded 
from  an  estimate  of  remedial  expenditure,  because  it  is  not  an  external  factor  acting  on  the 
process  but  an  element  of  the  latter's  normal  or  autonomous  mechanism.     As  state  enter- 
prise expands,  this  must  become  increasingly  important,  but  it  is  so  even  with  many 
expenditures  which  do  not  issue  in  monetary  returns  at  all  or  in  returns  which  are  com- 


1002  BUSINESS  CYCLES 

merely  note  that  net  national  income  increased  by  about  8.6  billions  in 
1934  and  by  about  5.2  billions  in  1935  (and  by  8.81  billions  in  1936),  which 
compares  with  federal  net  income-generating  expenditure  of  1,856 
millions  for  1933,  of  3,238  millions  in  1934,  and  of  3,154  millions  in  1935 
(and  4,025  millions  in  1936). 

As  measured  by  those  and  other  figures — for  instance,  of  employment 
(see  below) — effects  may  well  seem  surprisingly  small.  They  have,  in 
fact,  been  felt  to  be  so  even  by  those  economists  who  simply  attribute  the 
whole  of  the  observed  increase  in  national  income  to  federal  income 
generation.  But  obviously  we  cannot  do  this,  for  to  hold  that  income 
generation  alone  has  been  responsible  for  that  increase  involves  either 
circular  reasoning  or  else  the  theory  that  in  the  absence  of  it  the 
economic  process  would  have  gone  on  shrinking  or  would  have  dragged 
along  indefinitely  at  the  minimum  level.2  There  is  no  warrant  for  believ- 
ing this.  On  the  contrary,  there  is,  as  we  have  seen,  reason  to  believe 
that  there  would  have  been  recovery  in  any  case — a  recovery  strong 
enough  to  produce  by  itself  most  of  the  increases,  especially  in  output, 
that  actually  occurred  and  more.  It  follows  that  by  taking  account  of 
the  cyclical  phases  on  which  federal  income  generation  impinged,  our 
expectations  as  to  its  effects  can  be  only  raised  and  not  lowered.  We  will 
restate  these  expectations  under  four  heads. 

First,  government  expenditure  will  improve  any  business  situation, 
even  if  it  increase  the  national  income  of  the  year  only  by  the  amount 
spent  or  by  less  than  that  or  even,  in  a  limiting  case,  by  nothing  at  all, 
through  helping  the  public  to  build  up  depleted  balances  and  to  repay 
debts.  If  firms  repay  bank  loans  by  means  of  money  which  the  spending 
government  raised  from  existing  but  idle  deposits,  improvement  may  even 
be  accompanied  by  a  decline  in  total  or  total  demand  deposits  which, 
as  we  have  seen  before,  may  thus  be  a  favorable  symptom.3  As  a  rule, 

mercially  adequate:  public  works  using  means  of  production  that  would  otherwise  go  to 
waste  may  be  said  to  have  been  carried  out  at  no  social  cost  whatever. 

1  Professor  Kuznets'  figures  end  with  1935.     Hence,  the  figure  for  1936,  taken  from  the 
Department  of  Commerce  series,  is  not  strictly  comparable  with  the  preceding  ones. 

2  An  assumption  of  this  kind  underlies  the  reasoning  of  Professors  Colm  and  Lehmann 
in  their  otherwise  excellent  study  of  Public  Spending  and  Recovery  in  the  United  States, 
Social  Research,  May  1936.     Dr.  Currie  (paper  read  at  the  meeting  of  the  American  Eco- 
nomic Association,  Dec.  30,  1937)  argues  that  business  deposits  increased  strongly  from 
1933  to  1935,  while  business  borrowing  declined,  that  this  increase  can  only  have  been  due 
to  the  net  income-increasing  disbursements  by  government,  and  that  hence  public  spend- 
ing must  have  been  the  initiating  force  of  recovery.     This  is  a  clear  non  sequitur,  unless 
we  make  recovery  synonymous  with  increase  in  deposits  and  assume  quite  gratuitously 
that  the  decline  in  borrowing  was  completely  independent  of  those  disbursments.     See 
the  concluding  paragraphs  of  this  subsection. 

8  Of  course,  that  symptom  is  in  this  country  not  visible  in  any  deposit  series.  The 
behavior  of  loans,  however,  suggests  that  this  component  was  not  entirely  absent. 


THE  WORLD  CRISIS  AND  AFTER  1003 

only  a  moderate  amount  of  unfreezing  is  left  for  the  early  stages  of  the 
recovery  phase.  But  owing  to  the  extent  of  the  preceding  catastrophe 
and  of  the  continuing  state  of  overindebtedness,  it  is  reasonable  to  assume 
that  in  this  case  "consolidation"  was  one  of  the  major  remedial  effects  of 
the  spending  policy.1  Though  acting  in  the  monetary  sphere,  it  would, 
however,  primarily  show  outside  of  it. 

Second,  there  were  what  we  will  call  the  direct  results  of  handling 
the  government's  money  as  far  as  it  was  not  absorbed  by  the  replenishing 
of  balances  and  the  repayment  of  debts:  the  unemployed  man  spending 
his  dole,  the  man  who  has  been  reemployed  in  order  to  fill  a  government 
order  spending  his  wages,  at  the  retailer's  shop,  the  retailer  thereupon 
placing  additional  orders,  and  so  on.  Separate  evaluation  of  this  effect, 
which  was  no  doubt  considerable,  is  impossible  in  the  present  state  of  our 
information,  since  among  other  things  we  do  not  know  the  value  of  that 
income  efficiency  of  money  (Chap.  XI,  Sec.  A)  by  which  the  relevant 
part  of  government-created  income  would  have  to  be  multiplied.2 

Third,  firms — particularly  in  recovery — will  react  not  only  directly 
to  government  orders  or  to  purchases  by  the  first  recipients  of  government 
funds  but  also  indirectly  by  expanding  operations  in  anticipation  of  those 
orders  or  purchases  and  by  otherwise  "magnifying"  the  immediate 
effects  of  government  disbursements.  It  should  be  observed,  however, 
that  under  the  conditions  prevailing  in  the  early  stages  of  recovery 
new  investment  cannot  be  expected  to  be  much  in  evidence.  If  in  this 
instance  it  had  been,  this  would  before  1935  have  introduced  an  entirely 
abnormal  feature.  This  is  not,  of  course,  to  deny  that  stimulation  of 
investment  is  entitled  to  a  prominent  place  in  a  general  theory  of  govern- 
mental income  generation.  In  and  after  1935  some  such  stimulation 
may  have  been  present.  In  depression  and  recovery,  however,  it  is 

1  If  we  call  that  effect  remedial — and  the  same  term  could  be  applied  to  the  effects  to  be 
noticed  under  the  remaining  heads — we  do  not  thereby  "justify"  the  spending  policy. 
A  drug  may  be  "remedial"  with  respect  to  headache  but  "injurious"  to  the  heart.     The 
writer  entertains  no  doubt  but  that  public  income  generation  outside  of  "deep"  depression 
(roughly  from  the  middle  of  1930  to  the  middle  of  1931)  impairs  the  efficiency  of  the 
capitalist  process  for  reasons  that  should  be  familiar  by  now.     They  apply  especially  to 
public  income  generation  during  the  later  stages  of  recovery  and  the  first  stages  of  pros- 
perity.    The  difference  in  the  effects  of  public  and  entrepreneurial  expenditure,  the  latter 
involving,  the  former  not  involving,  change  in  production  functions,  should  particularly 
be  borne  in  mind. 

2  We  might,  of  course,  figure  out  average  income  "velocity,"  i.e.,  efficiency  times  rate 
of  spending,  in  the  usual  way  by  dividing  total  income  by  deposits  plus  money  in  hand-to- 
hand  circulation.     But  it  will  be  seen  on  reflection  that  we  cannot  arrive  at  the  effect  of 
federal  income-generating  expenditure  by  multiplying  the  result  into  the  amount  of  that 
expenditure.     Doing  this  would  precisely  involve  the  circular  reasoning  adverted  to  above. 
It  is  not  held  that  the  problem  is  insoluble,  but  its  solution  involves  a  formidable  research 
program. 


1004  BUSINESS  CYCLES 

current  operations  in  and  near  the  consumers'  sphere  that  need  to  be 
and,  as  a  matter  of  fact,  have  been  stimulated. 

Fourth,  irrespectively  of  reactions  of  the  type  noticed  under  the  pre- 
ceding heading,  there  will  be  ulterior  effects  on  economic  activity.  The 
relief  in  the  debt  structure,  the  steadying  of  prices,  the  improvement 
in  the  sectors  immediately  affected  by  government  disbursements,  the 
general  feeling  that  a  floor  is  being  provided  will  remove  inhibitions  and 
invite  advance  all  round.  This  class  of  effects  should  have  been  particu- 
larly strong  in  a  situation  in  which  not  only  the  stage  was  set  for  recovery 
but  in  which  a  recovery  that  had  already  begun  had  been  interrupted 
by  an  experience  so  trying  to  business  nerves  as  an  epidemic  among 
banks.  While  there  is  thus  no  intention  on  the  part  of  the  writer  to 
deny  the  reality  of  these  effects,  protest  must  be  entered  against  the 
practice  of  some  economists  in  making  an  uncritical  use  of  them,  which 
amounts  to  begging  the  question. 

It  should  be  added  that  federal  income  generation  must  also  have 
given  an  impulse  to  consumers'  credit  by  making,  directly  and  indirectly, 
many  households  "credit  worthy"  which  had  previously  ceased  to  be  so. 
1934,  in  fact,  displays  much  higher  figures  than  1933  for  the  credits 
outstanding  of  intermediary  and  cash-lending  agencies  and  also  for 
the  receivables  of  retail  merchants.  This  was  much  facilitated  by  the 
fact  that  this  type  of  financing  afforded  at  that  time  the  most  obvious 
chance  for  banks  to  respond  to  the  incessant  appeals  to  "liberalize" 
lending.  Since  the  above  analysis  applies,  with  but  little  modification, 
to  income  generation  by  consumers'  credit  as  well  as  to  income  generation 
by  government  expenditure,  it  is  easy  to  realize  that  the  former  reinforced 
the  effects  of  the  latter. 

This  analysis  evidently  harbors  no  tendency  to  underrate  the  poten- 
tialities of  pump  priming  which,  as  has  been  pointed  out  before  and  as 
the  war  experience  shows,  may  even  if  sufficiently  persisted  in  turn 
depression  into  a  state  displaying  all  surface  characteristics  of  prosperity. 
It  will  be  observed  that  in  some  points  we  do  not  share  the  disappointment 
felt  alike  by  advocates  of  governmental  income  generation  and  by  its 
opponents.  We  were  not,  for  instance,  disappointed  on  the  ground 
that  private  investment  was  not  more  strongly  stimulated  in  1934.  Nor 
did  we  consider  that  application  of  government-created  funds  to  the 
replenishing  of  balances  and  to  the  repayment  of  debt  constituted  pro 
tanto  defeat  of  the  spending  policy.  Observed  results  were,  nevertheless, 
no  better  than  they  could  have  been  expected  to  be  had  that  policy  been 
the  only  component  to  act.  Since  we  are  not  at  liberty  to  disregard 
the  other  component,  which  also  was  adequate  to  produce  those  results, 
we  are  driven  to  the  conclusion — to  be  verified  in  the  next  section — that 
other  factors  weakened  the  combined  effects  of  both. 


THE  WORLD  CRISIS  AND  AFTER  1005 

Since  so  many  economists  accept  the  quantitative  adequacy  of  the 
injection  of  purchasing  power  to  produce  observed  results  as  an  ipso  facto 
proof  that  there  cannot  have  been  any  contribution  from  the  economic 
process  itself — or  that  there  was  a  negative  one — it  will  be  well  to  retrace 
our  steps,  in  order  to  state  explicitly  the  case  for  alternative  possibilities. 
One  has  been  noticed  under  our  first  heading:  that  part  of  net  income- 
generating  expenditure  which  increased  national  income  only  by  its 
own  amount  or  less  will,  as  long  as  we  do  not  know  its  amount,  obviously 
be  credited  with  more  than  its  share  if  expansion  unconnected  with  it 
should  occur  at  the  same  time — financed,  let  us  say,  from  existing  depos- 
its— and  if,  on  the  strength  of  a  plausible  quantitative  relation  between 
total  income-generating  expenditure  and  total  increase  in  national 
income,  the  one  be  connected  with  the  other.  The  state  of  overindebted- 
ness,  on  the  one  hand,  and  the  cyclical  phase  on  the  other,  combine  to 
make  it  a  practical  certainty  that  this  coincidence  was  of  some  importance 
in  shaping  the  statistical  picture. 

Again,  if  government  funds  swell  business  deposits,  as  of  course  they 
did,  they  will  then  finance  the  subsequent  transactions  of  the  recipient 
firms,  whether  those  transactions  are  induced  by  the  act  of  expenditure 
or  not.  It  does  not  follow  that  every  expansion  of  operations  by  these 
firms  must,  therefore,  be  causally  related  to  those  receipts  and  that  with- 
out them  they  would  not  have  expanded  operations  at  all.  In  this  case 
it  is  indeed  still  more  obvious  than  it  is  in  the  other  how  government 
expenditure  propelled  recovery.  But  since  resumption  or  expansion 
is,  for  that  cyclical  juncture,  independently  motivated,  it  will  not  do 
to  attribute  all  that  was  financed  by  money  originally  inserted  by 
government  disbursements  to  the  impulse  imparted  by  them:  it  can- 
not be  inferred  that  government  carried  business  as  Aeneas  carried 
Anchises.  To  some  extent  at  least,  financing  by  receipts  simply  replaced 
financing  by  borrowing.  As  we  shall  see,  there  is  ample  evidence  for  this. 

It  should  not  be  replied  that  this  is  a  case  of  speculation  about  possi- 
bilities vs.  hard  statistical  fact.  It  is  a  case  of  common  sense  vs.  a  type 
of  monetary  theories.  What  we  see  is  the  income-generating  expenditure 
and  certain  developments.  The  relation  between  them  we  do  not  see. 
Our  interpretation  of  it  is  not  more  but  less  hypothetical  or  speculative 
than  the  one  which  exclusively  relies  on  mechanical  relations  between  the 
two,  because  it  assumes  and  asserts  much  less. 

Finally,  it  should  be  observed  that  no  account  has  been  taken  of  the 
possibility  that  government  spending  might  have  interfered  with  business 
expansion  (excepting,  of  course,  the  expansion  of  bank  loans).  Some 
of  the  arguments  adduced  for  this  possibility  fully  merit  the  shrugging 
of  shoulders  with  which  they  are  usually  met,  for  instance  the  argument 
that  the  unbalanced  budget  destroyed  confidence.  Others  do  not. 


1006  BUSINESS  CYCLES 

But  the  net  effect  of  the  spending  policy,  taken  by  itself  and  considered 
only  with  reference  to  the  general  complexion  of  short-run  business  situations^ 
seems  to  the  writer  to  have  been  so  clearly  positive  as  to  justify  him  in 
disregarding,  for  the  purpose  in  hand,  any  possibilities  of  that  kind.1 

6.  We  will  now  glance  at  the  statistical  picture.  It  should  be  recalled 
that  we  need  not  lean  too  much  on  "inflationary"  anticipations  or  on 
speculation  in  general  when  explaining  the  boom  of  1933,  which  culmi- 
nated in  July.  Nor  need  we  call  it  simply  an  affair  of  restocking  (inven- 
tory boom).  Though  all  these  and  other  factors — such  as  attempts  to 
profit  from  purchases  or  production  not  yet  burdened  by  the  extra  costs 
that  were  to  be  expected  from  AAA  and  NRA — no  doubt  materially 
helped,  that  boom  is  fundamentally  understandable  as  a  belated  and 
hence  more  violent  reaction  to  the  ravages  of  the  spiral,  the  intensity 
of  which  it  may  reasonably  be  said  to  measure.  Nor  is  it  necessary 
to  emphasize  the  element  of  reaction  to  excess  which  undoubtedly 
intensified  the  sharp  slump — it  was  sharper  than  almost  any  that  had 
preceded — in  the  third,  and  the  almost  level  movements  of  the  fourth 
quarter:  a  relapse  was  quite  within  ordinary  regularity  (Kitchin  depres- 
sion). But  public  spending  no  doubt  shortened  the  relapse  and  accentu- 
ated the  upswing — which  came  in  December — as  well  as  the  strong 
expansion  in  the  first  half  of  1934.  There  was  but  moderate  relaxation 
in  the  third  quarter  of  1934,  and  expansion  resumed  in  the  fourth 
quarter.  It  went  on  at  a  much  increased  rate  in  the  first  quarter  of  1935. 
Outside  debits  (141  cities;  Federal  Reserve  Board)  express  the  funda- 
mental contour  very  well  from  month  to  month.  For  the  years  in 
question,  monthly  averages  (which  were  at  a  maximum  of  27.66  billions 
in  1929;  at  23  billions  for  1930;  at  12.87  billions  in  1932)  struck  their 
low  point  of  12.2  billions  in  1933  (as  we  know,  mainly  in  consequence  of 
the  banking  troubles)  and  rose  to  but  13.83  billions  for  1934  (to  15.85 
billions  for  1935).  In  the  absence  of  government  income  generation 
we  should  have  called  this  conforming  to  expectation.  As  it  was,  the 
smallness  of  the  increase  becomes  a  problem.  Our  solution  of  this 
problem  is  indicated  in  the  preceding  paragraphs. 

Profits  fell  from  the  fourth  quarter  of  1932  to  the  first  quarter  of  1933, 
if  allowance  be  made  for  seasonal  behavior,2  and  recovered  strongly  in 
the  second  quarter,  but  less  than  uncorrected  figures  suggest.  Then 
they  relapsed  again  but  made  a  showing  much  beyond  expectation 

1  C/.,  however,  the  last  footnote  but  one.     Neither  NRA  codes  nor  showers  of  money 
are  calculated  to  vitalize  the  capitalist  process. 

2  Cf.  W.  L.  Crum,  The  Course  of  Corporation  Profits,  Review  of  Economic  Statistics, 
Mar.  15,  1934.     The  writer  would,  more  positively  than  Professor  Crum,  assert  that  the 
true  low  point  of  corporation  profits  comes  in  the  second  quarter  in  1932.     Improvement 
for  the  greater  part  of  the  second  half  of  that  year  is,  in  any  case,  beyond  question,  at  least 
as  far  as  Professor  Crum's  selected  list  of  163  corporations  is  representative. 


THE  WORLD  CRISIS  AND  AFTER  1007 

from  our  model  in  1934,  the  quarterly  average  for  that  year  (Federal 
Reserve  Bank  of  New  York)  being  about  75  per  cent  above  that  of  1933 — 
roughly  speaking,  the  writer  would  attribute  three-quarters  of  this 
increase  to  government  spending  and  the  NRA.1  Stock  prices  copy  this 
contour  fairly  closely.  It  should  not  be  forgotten,  however,  that  they 
merely  continued  the  upswing  of  the  third  quarter  of  1932.  The  monthly 
average  of  new  corporate  capital  issues  (domestic;  Financial  and  Com- 
mercial Chronicle;  maximum  1929:  666.8  millions)  touched  its  low  of  13.8 
millions  in  1933  and  increased  to  but  14.8  millions  in  1934.  This  is  even 
less  than  we  should  have  expected  on  the  strength  of  the  proposition 
that  recovery  does  not  typically  or  necessarily  start  from  real  investment 
and  still  less  from  capital  issues.  Government  disbursements  in  part 
supplied  the  funds  which  it  would  otherwise  have  been  necessary  to  raise 
by  issues  or  to  borrow  from  banks.  Also,  the  proceeds  of  the  issues  of 
1929  were  still  largely  unused. 

Some  idea  of  the  extent  to  which  government  disbursements  replaced 
bank  credit  by  enabling  firms  to  finance  from  receipts  can  be  gathered 
from  the  behavior  of  All  Other  Loans  of  reporting  member  banks.  Those 
loans  did  not  increase,  as  we  have  seen,  in  the  incipient  recovery  of  1932 
— which  conforms  to  expectation — but  went  on  falling  and  ended  their 
downward  course  by  a  sharp  drop  in  the  first  quarter  of  1933.  Then 
they  did  rise  moderately  until  the  beginning  of  the  fourth  quarter,  when 
they  decreased  again.  So  far  there  was  nothing  abnormal.  But  they 
did  not  at  all  participate  in  the  upswing  of  the  first  half  of  1934,  while 
United  States  securities  held  increased  by  about  1  billion  in  the  first 
quarter  and  by  another  half  toward  the  end  of  the  second.  Increase  in 
the  third  was  weak  and  short-lived  and  more  than  compensated  by  the 
fall  to  the  end  of  the  year,  and  no  increase  accompanied  the  strong 
upswing  of  the  first  quarter  of  1935:  firms  did  not  go  to  banks  for  what 
they  got  from  government.  It  will  be  recalled  that  we  observed  the  same 
phenomenon  in  Germany. 

Outside  net  demand  deposits  had  increased  during  the  second  half 
of  1932.  After  the  understandable  slump  of  the  following  spring — which 
brought  them  to  a  low  that  is  spurious  in  our  sense — they  increased  more 
than  All  Other  Loans  to  July  1933,  mainly,  of  course,  because  of  the 
increase  in  members'  investment  incident  to  federal  reserve  banks' 
open-market  purchases.  Their  rate  of  turnover  also  increased  at  the 
same  time.  Then  they  fell  and  rose  along  with  outside  debits  until 
the  middle  of  1934,  investment  again  taking  the  place  of  loans  in 
creating  them.  But  they  continued  to  increase  after  that,  in  response 

1  Monthly  averages  of  cash  dividends  (The  New  York  Times;  maximum  1930:  $386.5 
million;  banks  excluded)  were  216.5  millions  in  1932,  181.5  millions  in  1933,  206  millions  in 
1934  (and  226.3  millions  for  1935). 


1008  BUSINESS  CYCLES 

to  investment  that  did  not  take  the  place  of  loans  which  would  otherwise 
have  emerged,  until  the  end  of  the  period:  government  by  selling  defici- 
ency bills — the  cash  deficit  of  1934  in  the  sense  of  the  Daily  Statements 
was  over  4.5  billions — acquired  deposits  which,  when  used,  produced 
other  deposits:  the  old  method  of  war  finance.  Transfer  of  surplus  funds 
to  centers  and  influx  of  gold — at  the  new  par — of  course,  also  swelled  net 
demand  deposits.1  There  is  no  problem  in  any  of  these  movements  or 
in  the  growth  of  excess  reserves. 

We  have  noticed  how  interest  rates  behaved  under  the  monetary 
regime  that  obtained,  and  will  only  add  that  bond  yields  (corporate 
issues;  the  combined  index  of  Standard  Statistics)  fell2  (monthly  average) 
from  6.27  per  cent  in  1932  to  5.92  per  cent  in  1933  and  4.86  per  cent  in 
1934,  and  were  4.78  per  cent  for  March  1935.  The  B.L.S.  index  of  whole- 
sale prices,  excluding  foods  and  farm  products,  rose  from  its  minimum  of 
65.3  in  April  1933,  to  not  more  than  a  monthly  average  of  78.4  for  19343 
(75  for  1931),  the  greater  part  of  the  rise  occurring  before  1933  was 
out,  i.e.,  before  the  spending  program  had  had  time  to  produce  its  full 
mechanical  effects.  This  is  within  the  limits  of  what  might  have  been 
expected  as  the  result  of  a  rebound  from  panic  lows.  Besides  the 
presence  of  underutilized  resources,  the  weight  of  incessantly  increased 
productive  efficiency  is  mainly  responsible  for  the  failure  of  prices  to 
respond  more  strongly  to  the  price-raising  policy4  of  the  administra- 
tion. The  latter  was  more  "successful"  with  respect  to  the  prices 
of  foods  and  farm  products.  Prices  received  by  farmers  increased 
by  84  per  cent  (index  of  U.  S.  Department  of  Agriculture)  between 
March  1933  and  December  1934.  The  B.L.S.  index  for  farm  products 
rose,  however,  from  40.9  for  February  1933,  to  78.3  at  the  end  of  our 
period,  viz.,  for  March  1935,  i.e.,  by  over  91  per  cent  of  its  minimum. 
But  the  reader  will  realize  that  it  would  be  easy  to  choose  dates  in 
such  a  way  as  to  get  almost  exact  correspondence  between  that  increase 
and  the  decrease  in  the  gold  content  of  the  dollar — he  will  also  realize 
how  perfectly  meaningless  that  would  be. 

1  Through  August  23,  1935,  the  term  net  demand  deposits  retains  its  old  meaning.     Total 
deposits  of  reporting  member  banks  increased  by  about  3  billions  during  the  year. 

2  Along  with  a  fall  in  stock  prices,  bond  yields  increased  somewhat  in  the  third  quarter  of 
1934,  while  there  was  a  temporary  outflow  of  gold.     This  was  a  not  unnatural  concomitant 
of  deficit  financiering,  but  on  the  whole  such  effects  were  successfully  avoided.     However, 
bond  yields  were  not  low  enough  obviously  to  negative  the  idea  of  the  presence  of  distrust 
in  the  currency.     The  fact  that  really  cheap  rates  were  confined  to  the  open  market  (in  our 
sense)  does  point  in  that  direction. 

3  The  Fairchild  retail  price  index,  which  also  excludes  foods  (December  1930  =  100) 
and  also  has  a  minimum  (69.4)  for  April  1933,  rose  to  a  monthly  average  of  88.3  for  1934. 

*  Prices  in  full- weight  dollars  fell  until  November  1933  and  did  not  really  start  to  rise 
until  May  1934.     But  there  is  no  need  of  going  into  the  various  aspects  of  this. 


THE  WORLD  CRISIS  AND  AFTER  1009 

Average  hourly  money  wage  rates  behaved  very  differently  in  differ- 
ent sectors — in  anthracite  coal  mining  they  increased  very  little,  in 
bituminous  coal  very  much,  for  instance,  which  again  illustrates  the  lack 
of  realism  incident  to  speaking  of  a  wage  level.  On  the  whole,  however, 
these  different  rates  of  increase  seem  to  have  worked  in  the  direction  of 
a  more  balanced  wage  structure.  For  manufacturing  industries  as  a 
whole  ("all  wage  earners":  25  industries)  an  increase  of  over  22  per  cent 
occurred  between  June  (the  minimum)  and  December  1933.  After 
that  there  was  no  relapse — though  there  was  one  in  the  wages  paid  by  the 
wholesale  and  retail  trade — but  an  increase  at  a  much  slower  rate, 
which  was,  nevertheless,  sufficient  to  produce  an  increase  in  real  rates  of 
about  6  per  cent  by  March  1935.  The  annual  average  of  the  hourly 
rates  of  skilled  and  semiskilled  labor  in  manufacturing  industries  (National 
Industrial  Conference  Board)  was  55  cents  for  1933  and  a  little  over  64 
in  1934. l  This  is  clearly  contrary  to  expectation  from  our  model. 
Excepting,  perhaps,  a  fraction  of  the  increase  in  1933  which  might  have 
resulted  in  any  case  from  the  general  reaction  to  panic  lows,  it  must  be 
attributed  to  government  policies  which  are  the  only  available  factors  to 
explain  how  the  upward  shift  of  the  demand  " curve*'  for  labor  could 
have  produced  such  rates  in  spite  of  the  prevailing  unemployment.  No 
comments  about  consequences  need  be  added  to  what  has  been  said  on 
the  subject  before  (see  especially  above,  sub  3). 

Both  factory  pay  rolls  and  employment  started  on  their  upward  course 
before  monetary  wage  rates.  In  fact,  it  is  important  to  recall  that  they 
began  to  increase  in  the  third  quarter  of  1932  and  that  the  increase  which 
occurred  from  March  to  September  1933  was  but  a  continuation  and  not  a 
break.  Monthly  averages  (combined  index  of  the  Department  of  Labor) 
of  factory  pay  rolls  were  for  1933  about  6  per  cent  above  1932,  for  1934 
nearly  21  per  cent  above  1933,  and  for  1935  a  little  over  13  per  cent  above 
1934,  while  the  corresponding  increases  in  employment  averages  were  10, 
13,  and  a  little  over  4  per  cent,  the  latter  figure  being  suggestive  of 
inhibitions.  Particulars  of  incidence  of  unemployment  or  reemployment 
were  as  we  should  expect  and  need  not  be  discussed  here. 

Output  of  manufacturing  and  mining,  as  measured  by  the  Federal 
Reserve  Board's  index  and  as  reflected  in  production  of  electric  power — 
which,  even  if  corrected  for  "trend,"  increased  by  19  per  cent  from  March 
to  July — and  in  carloadings,  also  behaved  according  to  expectation  from 
our  model  modified  by  the  political  factor.  Building  contracts,  for 
example,  were,  of  course,  affected  by  the  public  works  program  and  would 

1  Per  capita  weekly  earnings  in  manufacturing  began  to  rise  earlier — already  in  the 
spring — and  then  increased  with  fluctuations  to  the  third  quarter  of  1986.  Of  course,  they 
also  increased  in  real  terms,  hours  worked  per  week  rising  somewhat  above  their  1932 
minimum  of  34.8. 


1010  BUSINESS  CYCLES 

not  otherwise  have  increased  much  in  1933 — private  building  was  indeed 
below  expectation,  if  anything.  The  sharp  increase  in  steel-ingot  produc- 
tion in  the  second  quarter  of  1933  was  from  a  very  low  level,  but  even  so 
surprisingly  great — over  300  per  cent  in  4  months,  thereby  reaching  the 
1923  to  1925  average — so  that  the  subsequent  reaction  of  over  50  per  cent 
was  very  understandable.  Neither  its  rise  nor  its  fall,  however,  meant 
what  they  used  to  mean  of  old,  when  steel  satisfactorily  represented 
equipment.  But  it  should  be  observed  that  from  any  reserve  capacity 
which  may  have  existed  in  1929  and  which  was  obviously  greatly  increased 
by  the  shrinkage  of  the  subsequent  years,  a  substantial  deduction  must 
be  made  on  the  score  of  wear  and  tear — some  of  it  independent  of  degree 
of  utilization — and  of  obsolescence. 

Although  machine  tool  orders  rose  to  a  fairly  high  level  at  the  end  of 
the  period,  steel  output  was,  during  1933  and  1934,  primarily  associated 
with  durable  consumers'  goods.  Output  of  automobiles  nearly  reached 
3  million  units  (passenger  cars  and  trucks)  in  1934.1  Refrigerators, 
air-conditioning  installations,2  and  other  members  of  the  class  of  "depres- 
sionless  industries,"  such  as  gasoline,  cigarettes,  rayon,  and  some  chemi- 
cals, showed  considerable  gains. 

In  the  behavior  of  output  three  points  must  be  noticed.  First,  short 
fluctuations  do  not  quite  correspond  to  our  idea  of  what  they  should  have 
been.  The  increase  of  66  per  cent  from  March  to  July  1933,  which  carried 
it  to  the  1923  to  1925  average  and  the  subsequent  decline  are  at  least 
timed  according  to  expectation.  But  the  increase  in  the  first  half  of 
1934  came  6  months  earlier  than  we  should  have  expected — relapse  in 
fact  followed — and  for  the  decline  and  stagnation  in  the  first  half  of  1935 
we  have  only  political  influences  ( ?)  to  offer.  Second,  the  index  did  not, 
even  before  the  decline  early  in  1935  and  excepting  the  1933  peak,  reach 
the  level  of  1925-1926,  the  preceding  neighborhood  of  equilibrium,  while 
it  should  have  surpassed  it.  We  attribute  this  to  the  severity  of  the 
depression  and  may  recall  that  the  recovery  after  1873  also  was  not  satis- 
factory in  every  way.  Nevertheless,  effects  of  the  NRA  policy  may  be 
partly  responsible.  Third,  we  must  supplement  the  case  for  an  adverse 
effect  of  the  increase  in  wage  rates  on  employment  by  noting  the  striking 
difference  between  the  latter  and  the  corresponding  output.  To  that 
66  per  cent  increase  in  output  corresponds  an  increase  of  only  33  per  cent 
in  factory  employment.  This  is  in  part  due  to  underutilization  of  the 
working  force  employed  around  the  bottom  of  the  depression — average 

1  This  was  not  simply  "rebound."     Much  had  changed  in  the  industry  during  depres- 
sion years,  and  part  of  that  increase  must  be  attributed  to  its  own  impulse.     There  is  a 
kernel  of  truth  in  the  exaggeration  that  "motors  led  us  out  of  depression." 

2  The  latter,  however,  were  not  significant  quantitatively.     They  are  more  properly 
considered  as  an  innovation  of  the  following  Juglar. 


THE  WORLD  CRISIS  AND  AFTER 


1011 


hours  worked  per  week,  which  may  be  taken  as  an  indicator,  in  fact, 
increased  by  nearly  5  per  cent  from  1932  to  1933 — and  in  part  simply  the 
consequence  of  labor-saving  rationalization  which  had  been  going  on 
through  the  depression.  But  it  is  impossible  to  overlook  the  premium 
that  the  wage  policy  set  on  this  rationalization.1 

To  sum  up:  The  reader  will  have  no  difficulty  in  listing  the  symptoms 
by  which  the  effects  of  governmental  income  generation  unmistakably 
show.  In  spite  of  them,  however,  the  statistical  picture  presented  does 
not  differ  fundamentally  from  what  we  should  have  expected  to  see  in 
the  absence  of  that  factor.  Since,  as  has  been  pointed  out  in  5,  it  would 
have  been  possible  for  such  a  picture  to  emerge,  public  expenditure  not- 
withstanding, under  the  sole  influence  of  the  normal  recovery  process, 
the  conclusion  seems  to  suggest  itself  that — barring  minor  deviations 
caused  by  it,  such  as  we  observed  in  the  behavior  of  deposits — that 
expenditure  took  no  effect.  This  conclusion  we  do  not  draw.  If  we 
did,  we  should,  in  fact,  be  committing  the  very  same  error  which  those 
economists  commit  who  simply  attribute  to  government  spending  every- 
thing that  happened.  But  we  do  draw  two  other  conclusions:  first,  that 
attributing  all  observed  developments  to  the  normal  recovery  process 
would,  though  wrong,  not  be  more  so — or  more  "speculative" — than 
the  opposite  opinion;  and  that  the  prima-facie  impression  with  which  we 
started  in  5  is  misleading. 

G.  The  Disappointing  Juglar. — If  past  experience  be  a  guide  and  our 
schema  a  roughly  correct  expression  of  it,  then  the  rise  of  a  new  Juglar — 
the  fifth  of  the  Kondratieff — was  due  for  the  spring  of  1935,  however 
little  meaning  we  may  attach  to  the  precise  date  yielded  by  our  experi- 
mental count  (beginning  of  April).  Though,  as  the  reader  knows  and 
as  we  shall  presently  sec,  facts  did  not  entirely  fail,  for  a  time  just  about 
equal  to  the  duration  of  an  average  Juglar  prosperity,  to  bear  out  the 
expectation  which  would  follow  from  that,  the  difference  is  great  indeed 


CHANGES   IN    HOURS,    EARNINGS,    PRODUCTION,   AND    PRICES,  MANUFACTURING   INDUSTRIES,  BY   MONTHS 

103£-1935 
(Source:  Monthly  Labor  Review) 


Year, 
monthly  average 

Index  of 
total 
man-hours 

Index  of 
average 
hourly 
earnings 

Index  of 
production 

Index  of 
average 
output  per 
man-hour 

Index  of 
labor  cost 
per  unit 
of  output 

Wholesale 
prices  for  all 
commodities 
other  than 
farm  products 

1982 

100  0 

100  0 

100  0 

100.0 

100  0 

100  0 

1933 

107  3 

98  1 

119  0 

111  0 

88  0 

101  0 

1934 

114  4 

116  0 

124  0 

108  0 

108  0 

112.6 

1935 

126  2 

120  1 

143.0 

118  0 

107.0 

117.4 

1012  BUSINESS  CYCLES 

between  such  upswing  as  there  was  and  what  happened  in  the  last 
comparable  instance,  1879-1880.  Conditions  external  to  this  country 
which  then  produced  an  agricultural  boom  may  no  doubt  be  invoked  in 
partial  explanation.  But  government  policy  largely  did  for  agriculture 
now  what  European  demand  had  done  for  it  then,  and  if  we  take  account 
of  government  expenditure  in  general,  agrarian  and  other,  the  picture 
becomes  still  more  disappointing.  This  was,  in  fact,  universally  felt. 
People  never  spoke  of  more  than  recovery  and  an  unsatisfactory  one  at 
that.  It  would  not,  however,  help  us  much  if  we  did  the* same;  for  the 
real  trouble  with  expectation  from  our  model  is  not  in  the  weakness  of 
that  "prosperity"  but  in  the  fact  that  it  was  followed,  instead  of  by  a 
recession  in  our  sense,  by  a  break  which  landed  the  system,  within  a  few 
months  and  at  a  rate  surpassing  everything  witnessed  during  the  years 
from  1930  to  1932,  in  a  state  displaying  all  the  phenomena  of  deep 
depression.  This  would  be  still  more  unexpected — from  the  standpoint 
of  our  model  as  such — as  a  sequel  to  a  recovery  than  it  is  as  a  sequel  to  a 
prosperity  phase.  Does  it  mean  that  the  capitalist  process  has  spent  its 
force,  that  private  investment  opportunity  has  vanished  to  the  point 
of  making  it  dependent  on  government  expenditure  for  motive  power  or 
in  such  a  way  that  the  system  must  collapse  as  soon  as  government  expen- 
diture is  withdrawn,  like  one  of  those  children's  balloons  that  shrivel  as 
soon  as  one  ceases  to  blow  air  into  them? 

Reasons  for  believing  that  this  is  unlikely  have  been  offered  in  the 
introduction  to  the  preceding  section.  But  there  seems  to  be  evidence 
for  it  of  an  almost  experimental  nature.  The  figures  of  federal  income- 
generating  expenditure  for  1935  and  1936  have  been  mentioned.1  Again 
it  was,  prima  facie,  quite  sufficient  to  justify  the  statement  that  it 
induced,  directly  and  indirectly,  such — this  time — prosperity  as  there 
was  during  those  two  years,  and  if  we  take  into  account  deferred  effects, 
also  during  the  first  half  of  1937.  Though  there  were  developments  that 
seem  to  be  beyond  the  range  of  its  consequences,  it  might  still  be  said 
without  absurdity  that  the  spending  policy  then  "took  effect  at  last." 
After  that  it  was  discontinued.  On  cash  account  the  Treasury  got  out  of 
the  red.2  A  slump  ensued  in  due  course.  Moreover,  the  writer  enter- 
tains no  doubt  not  only  that  that  slump  will  give  way  to  recovery  as  the 
new  spending  program  within  the  4  billion  deficit  budgeted  for  1938 

1  The  veterans*  cash  bonus  of  1,900  millions  became  payable  from  June  15,  1936.     Half 
of  it  enters  into  Mr.  Currie's  figures,  being  distributed,  as  seems  reasonable,  over  the  month 
of  payment,  the  two  months  preceding,  and  the  three  months  following.     The  writer  would 
have  distributed  the  whole  of  it  in  some  such  way.     As  a  matter  of  fact,  about  1,200  millions 
were  promptly  cashed,  and  chain-  and  department-store  sales  responded  immediately. 

2  The  figure  of  income-generating  expenditure  for  1987,  $900  million,  must  be  interpreted 
with  due  regard  to  the  fact  that  this  sum  was  almost  entirely  spent  in  the  first  seven 
months  of  the  year. 


THE  WORLD  CRISIS  AND  AFTER  1013 

unfolds  during  the  fall  of  1938,  but  also  that  tapering  off  will  again  be 
attended  by  the  symptoms  of — according  to  the  way  in  which  it  is 
effected — recession  or  depression.1  This  should  make  us  both  envious 
and  thankful:  envious  because  fellow  economists  will  be  able  to  enjoy 
so  delightful  a  verification  of  their  views,  thankful  because  in  other  fields 
— medicine,  for  instance — people  do  not  reason  like  that,  or  else  we  should 
all  of  us  be  morphinists  by  now. 

1.  In  order  to  see  more  precisely  what  there  is  to  explain,  we  will 
begin  by  a  survey  of  time-series  contours.  1935  was  the  third  year  to 
show  almost  consistently  higher  levels  of  annual  figures  than  its  prede- 
cessor. Monthly  figures,  however,  were  not  consistently  higher.  The 
weekly  operating  rate  of  the  steel  industry  may  serve  as  an  example. 
Excepting  the  first  two  months,  it  was  at  or  below  the  1934  level  until 
nearly  the  end  of  June;  only  in  the  second  half  of  the  year  did  it  rise 
above  that.  By  end  of  September  it  was  50  per  cent — in  itself  sufficient 
proof  of  the  weakness  of  that  upswing,  though  90  per  cent  was  eventually 
reached  (March—April  1937).  Moreover,  there  were  other  irregularities, 
among  them  two  setbacks  in  which  leading  series  behaved  in  a  somewhat 
discordant  manner:  during  the  first  half  of  the  year  it  was  the  index  of 
production  of  manufactures  and  minerals  and  allied  indices,  such  as 
carloadings,  that  sagged,  while  outside  debits  steadily  increased;2  the 
little  relapse  in  the  autumn  shows  primarily  in  outside  debits,  while 
production  was  hardly  affected  at  all.  These  relapses  and  discordances 
were — though  on  a  very  small  scale — repeated  in  1936, 3  which  otherwise 

1  That  sentence  has  been  left  standing  as  it  was  written  in  July  or  August  1938.     It  may 
be  useful  to  add  the  following  comments  (May  1989). 

Marked  improvement  in  fact  showed  in  the  third  and  fourth  quarters  of  1938,  espe- 
cially in  the  indices  of  manufacturing  output,  construction,  carloadings,  department-store 
sales,  employment,  though,  conforming  to  expectation,  prices  continued  to  decline.  These 
facts  obviously  bear  out  the  first  part  of  the  statement  in  our  text. 

But  the  relapse  during  the  first  quarter  of  1939,  which  continued  during  April,  does 
not  illustrate  the  second  part.  There  was  indeed,  a  considerable  increase  in  Treasury 
deposits  during  February,  due  to  the  sale  of  savings  bonds  and  of  securities  issued  on 
behalf  of  various  public  credit  agencies,  which  for  a  few  weeks  raised  government  cash 
receipts  above  government  disbursements.  This,  however,  was  hardly  adequate  to  pro- 
duce the  observed  results,  and  beyond  this  there  was  nothing  but  talk.  We  are  still  within 
the  rising  tide  of  spending,  and  if  effects  do  not  show  more  visibly,  this  is  due,  apart  from 
the  presence  of  depressing  extrasystematic  factors,  to  the  cyclical  phase.  If  the  reader 
refer  to  our  schema  he  will  see  that,  barring  possible  reactions  to  the  abnormal  slump  of 
1937-1938,  there  is  little  reason  to  expect  from  the  mechanism  of  our  process  any  very 
strong  upturn  for  several  years  to  come. 

2  That  in  the  beginning  of  the  prosperity  phase  money  volume  of  transactions  should 
increase  more  than  physical  output  is  not  in  itself  an  irregularity;  but  output  fell,  especially 
in  steel,  cotton  and  silk  textiles,  and  bituminous  coal. 

3  But  recalling  our  experimental  count  we  shall  not,  as  we  must  for  1935,  consider  the 
occurrence  of  those  setbacks  as  abnormal. 


1014  BUSINESS  CYCLES 

is  the  year  of  the  strongest  and  most  nearly  uninterrupted  increase  all 
round.  At  its  end  the  1929  peak  of  the  Federal  Reserve  Board's  index 
of  production  of  manufacturing  and  mining  was  almost  reached  (with 
120  per  cent  of  the  1923  to  1925  average;  from  May  1935  to  May  1936 
the  index  increased  by  16  points,  and  the  value  in  current  dollars  of 
national  income  produced  (63.8  billions)  was  almost  80  per  cent  of  the 
1929  figure.1  Aggregate  profits  of  700  industrial  and  mercantile  com- 
panies (Federal  reserve  bank  of  New  York),  which  had  in  1935  been  about 
80  per  cent  above  1934,  further  increased  by  about  50  pe*r  cent  in  1936, 
aviation  heading  the  list  and  steel,  automobiles,  tires,  petroleum, 
chemicals  and  drugs,  machinery  and  tools  showing  up  particularly  well. 
Only  6.4  per  cent  of  those  companies  reported  net  losses. 

As  stated  above,  a  Juglar  recession  in  our  sense  was  due  for  1937 
(the  middle  of  August  in  the  experimental  count).  But  what  actually 
happened  was  very  different.  After  a  drop  in  January,  outside  debits 
recovered  through  May — not,  however,  to  the  figure  of  the  preceding 
December — then  hovered  on  a  horizontal  level  until  August — so  far  con- 
forming to  expectation — and  after  that  shrank  rapidly.  In  the  first 
half  of  1938  they  continued  to  decline — not  merely  in  function  of  falling 
prices — but  at  a  decreasing  rate.  By  June  they  seemed  to  have  reached 
an  even  level.  Output  of  manufacturing  and  mining  behaved  irregularly 
from  the  first,  and  more  irregularly  than  debits  in  the  second  half  of 
1937;  instead  of  the  increase  that  we  should  expect,  we  find  that  it 
declined  in  January  and  recovered  only  to  May,  after  which  it  fell  by 
about  one-third  until  the  middle  of  1938,2  at  a  rate  that  decreased  from 
November  until  a  floor  was  apparently  reached  in  June.  Durable 
goods,  especially  equipment  goods,  for  which  demand  temporarily 
ceased  altogether,  of  course,  suffered  most.3 

Profits  behaved  similarly  and,  hence,  until  the  last  quarter  of  1937, 
conform  better  to  expectation.  As  far  as  data  of  quarterly  reporting 

1  Weekly  operating  rate  of  steel  was  at  about  70  per  cent,  or  above  that,  from  the  middle 
of  April  on,  about  75  per  cent  from  the  middle  of  September  on.     It  touched  80  in 
December. 

2  Steel  and  automobiles  kept  up  well  during  the  first  8  months  of  1937,  while  cotton, 
wool,  leather,  and  other  lines  declined  already  during  the  first  half  of  the  year.     Then  the 
iron  and  steel  index  dropped  by  65  per  cent  and  the  automobile  index  by  over  50  per  cent 
within  4  months.     Production  of  electric  power,  which  had  increased  strongly  in  1935  and 
still  more  so  in  1936,  surpassed  the  latter  record  until  August  and  then  fell  precipitously. 
The  operating  rate  of  the  steel  industry,  90  per  cent  in  March  and  April  1937  (see  above) 
was  down  to  25  per  cent  at  the  beginning  of  January  1938.     It  began  to  rise  at  the  begin- 
ning of  June. 

3  Department-store  sales,  which  from  the  middle  of  1935  to  the  middle  of  1936  had  risen 
by  about  16  per  cent,  were  well  sustained,  and  even  at  the  end  of  the  year  not  much  below 
the  1936  peak;  rural  sales  did  still  better  and  were  higher  for  the  second  than  for  the  first 
half. 


THE  WORLD  CRISIS  AND  AFTER  1015 

concerns  allow  us  to  judge,  profits  were  larger  by  one-half  in  the  first 
quarter  of  1937,  and  lower  by  about  one-third  in  the  fourth,  than  they 
had  been  in  the  corresponding  quarters  of  1936.  For  the  year  as  a 
whole  they  were  higher  by  about  7  per  cent.1  Of  the  700  concerns 
9.6  per  cent  reported  net  loss.  Substantial  gains  on  1936  figures  were 
shown  by  steel,  railroad  equipment,  machinery,  agricultural  implements, 
electrical  equipment,  oil,  metal,  and  mining  including  copper  and  copper 
products,  the  automobile  industry  being  among  the  chief  mourners. 

Factory  employment  rose  and  fell  much  less  than  output  all  along. 
In  the  monthly  average  of  1935  it  was  but  little  over  4  per  cent  above  the 
level  of  1934,  then  increased  at  a  somewhat  greater  rate  in  1936  and 
reached  its  maximum,  a  little  over  the  1923  to  1925  average,  in  the 
second  quarter  of  1937.  From  July  to  December  it  fell  by  14  per  cent. 
But  it  was  still  84  per  cent  of  that  average  early  in  1938  and  79  per  cent 
for  April.  Subsequent  changes  in  most  lines  were  small,  but  decreases  in 
the  steel,  machinery,  motorcar,  and  men's  clothing  industries  still  further 
reduced  the  index  through  May.  Pay  rolls,  of  course,  rose — also  to 
about  the  1923  to  1925  average  in  the  first  half  of  1937— and  fell,  by  23 
per  cent  from  July  to  December,  more  than  employment.  But  the  rise 
above  and  the  fall  below  proportionality  reflect  not  only  advances  or 
reductions  in  rates  but  also  increased  or  decreased  employment  in  indus- 
tries paying  higher  than  average  wages  and  decrease  or  increase  of  part- 
time  employment,  here  and  there  even  overtime — but  since  this  was 
largely  due  to  the  reduction  in  hours,  it  was  but  another  form  of  increase 
in  rates — and  its  elimination.2 

Money  wage  rates  gained,  as  we  have  seen,  less  in  1935  than  they  had 
gained  in  1934.  Then  they  scored  further  gains.  Average  hourly  rates 
for  skilled  and  semiskilled  males  in  manufacturing  (64  cents  in  1934) 
were  about  66  cents  in  1935  and  about  69  cents  in  1936.  Unskilled  males 
did  worse,  and  women  suffered  a  small  loss  in  reaction  to  the  particularly 
strong  increase  which  had  previously  occurred  in  their  case.  In  1937  rates 
were  increased  by  another  10  per  cent,  reaching  a  level  substantially 
above  that  of  1929,  while  real  rates  then  surpassed  the  latter  more  than 
25  per  cent.3  Labor  cost  per  unit  of  product  also  rose.  Again,  it 

1  Total  national  income  was  $69  billion. 

2  As  always,  indices  of  retail  sales  and  of  total  wage  payments  moved  closely  together. 
For  example,  for  1929  to  1935  the  chart  at  the  bottom  of  the  last  page  of  the  Cleveland  Trust 
Company's  Business  Bulletin  for  Apr.  15,  1936. 

3  Real  rates  were,  after  their  rise  in  1933,  substantially  constant  during  1934, 1985,  and 
1936.     They  rose  by  about  6  per  cent  in  1937,  as  they  should  in  recession.    There  was 
further  increase  in  the  first  half  of  1938. 

According  to  the  National  Industrial  Conference  Board,  Wages,  Hours  and  Employ- 
ment in  the  United  States,  1914-1936,  Table  2,  average  wage  rate  of  factory  hands  was 
about  62  cents  in  1936  (working  hours  per  week  about  39).  This  compares  with  a  rate  of 


1016  BUSINESS  CYCLES 

follows  from  previous  argument  that  this  must  have  been  a  major  factor 
in  the  industrial  situation  which  it  tended  to  make  more  sensitive  to 
depressive  influences,  and  a  major  reason  why  unemployment,  in  spite 
of  a  substantial  decline  in  1936 — there  was  even  sporadic  shortness  of 
labor — remained  at  a  high  figure.  Besides  forcing  the  pace  of  labor- 
saving  rationalization,  it  may  in  spots  even  have  interfered  with  expan- 
sion of  output  while  prosperity  lasted.  Construction  may  serve  as  an 
example. 

The  strong  increase  in  contracts  awarded  which  we  observe  in  the 
fourth  quarter  of  1935  was  not  repeated  in  1936,  although  both  total  and 
privately  financed  construction  gained  considerably  over  the  year.  In 
1937  publicly  financed  construction  (a  little  over  1  billion)  declined  by 
over  15  per  cent,  but  privately  financed  contracts  for  building  and 
engineering  work  (about  1.8  billions  in  the  37  States)  showed  an  increase 
of  almost  40  per  cent  over  1936,  to  which  public  utilities  contributed 
considerably.  The  index  advanced  at  the  end  of  the  year  and  into 
January  1938.  *  In  February  there  was  a  sharp  fall  in  nonresidential 
construction — though  this  was  partly  made  up  for  in  March — and  pri- 
vately financed  contracts  for  the  quarter  were  30  per  cent  below  the  first 
quarter  of  1936.  But  even  before  the  slump,  the  showing  was  again, 
as  in  the  recovery,  below  what  we  should  expect  considering  possibilities 
(see  below)  and  monetary  conditions.  Building  costs,  which  rose  to  the 
1929  level  in  1936 — they  increased  further  by  nearly  20  per  cent  in  1937 — 
and  the  role  of  wage  rates  in  them  obviously  supply  the  explanation. 
However,  it  must  be  emphasized  once  more  that  such  increase  in  hourly 
rates  as  occurred  from  the  spring  of  1935  to  the  last  quarter  of  1936  is  not 
only  quite  within  expectation,  a  rise  of  money  wage  rates  being  a  normal 
element  of  prosperity,2  but  also  that  this  rise  cannot  be  thought  of  as  a 
cause  of  the  slump.  Not  even  the  total  rise  from  the  depression  low  can, 
except  as  an  element  of  a  complex  pattern  (see  below,  Sec.  G,  5,  c) — here, 
as  always,  we  must  beware  of  two- variable  arguments. 

2.  Conditions  of  extreme  monetary  ease  prevailed  throughout. 
Prime  commercial  paper,  for  instance,  was  at  %  of  1  per  cent  until  March 
1937.  Then  it  managed  to  climb  to  1  per  cent,  but  relapsed  to  the  old 
figure  by  June  1938.  Rates  on  customers'  loans,  which  on  May  31,  1935, 

25  cents  in  1914  (working  hours  per  week  about  51.5).  Monthly  average  of  the  board's 
cost-of-living  index  (1923  =  100)  was  61.3  for  1914  and  82  for  1936,  i.e.,  it  increased  by 
about  34  per  cent. 

1  This  was,  however,  partly  due  to  the  new  building  code  that  New  York  City  put  into 
force  at  the  end  of  January. 

2  It  cannot  be  objected  that  this  is  speaking  from  a  model  which  does  not  include  the 
possibility  of  abnormal  unemployment  in  prosperity.     It  does;  for  if  the  labor  market  be 
imperfect,  unemployment  will  not  prevent  a  rise  in  wage  rates. 


THE  WORLD  CRISIS  AND  AFTER  1017 

had  been  1.83,  fell  to  1.67,  and,  after  an  excursion  to  1.71,  to  1.63  by 
May  1938.  In  spite  of  the  heavy  treasury  financing,  the  yield  of  treasury 
bonds  declined,  after  a  slight  rise  in  the  third  quarter  of  1935,  to  a 
little  over  2.2 — a  development  which  was  accompanied  by  great  refunding 
operations  from  the  beginning  of  1935 — and  the  yield  of  AAA  corporate 
bonds  almost  without  interruption  to  4.5  at  the  end  of  1935  and  to  a 
little  over  3  per  cent  at  the  end  of  1936.  Then  there  was  an  increase 
through  April,  but  decline  to  new  lows  in  May  1937,  mortgage  rates 
keeping  again  a  much  higher  level  all  along.  This  behavior  of  short 
rates  and  yields  obviously  calls  for  explanation  other  than  can  be  gleaned 
from  our  model.  We  cannot  even  see  a  trace  of  the  expected  effect  in 
the  slight  increase  that  occurred  late  in  the  upswing,  since  that  resulted 
from  measures  of  monetary  management  (see  below).  But  the  irregu- 
larity should  not  be  exaggerated.  We  shall  see  it  in  its  true  proportions 
if  we  remember  the  behavior  of  English  rates  from  1873  to  1896. 

It  must  also  be  emphasized  again  that  the  financing  of  1929  still 
exerted  effects  which  should  also  go  some  way  toward  mitigating  surprise 
at  the  low  figures  of  domestic  corporate  issues.  The  monthly  average 
of  378  millions  in  1936  measures  well  up  to  the  1925  to  1929  average,  but 
most  of  those  issues  were  for  refunding.  Only  the  second  quarter  of 
1937  surpasses,  with  140  millions  per  month,  the  modest  figure  of  1931 
for  new  capital  issues.1  But  under  the  circumstances  full  time  at  machine 
and  machine  tool  shops  was  perfectly  compatible  with  that.  As  far  as 
this  goes,  the  common  saying  that  private  investment  did  not  really 
revive  during  the  upswing  must  be  modified.  Other  Loans  behaved,  in 
fact,  more  nearly  according  to  expectation  and  the  shift  in  bank  assets 
that  is  characteristic  of  prosperities  did  not  altogether  fail  to  show :  other 
loans  of  reporting  member  banks  (101  cities)  started  to  increase  from  the 
end  of  1935,  continued  to  do  so  up  to  the  last  quarter  of  1937,  and  then 
fell,  while  investments,  partly  at  least  in  connection  with  this,  fell  from 
about  the  middle  of  1936  and  began  to  increase  again  in  the  last  quarter 
of  1937  and  at  the  beginning  of  1938.  Variations  in  deposits  were  too 
much  under  the  influence  of  monetary  policy  and  government  action  in 
general  to  be  trusted  to  reflect  the  pulse  of  our  process,  except  perhaps 
in  the  second  half  of  1936  (see  below,  4) . 

All  that  need  be  said  about  stock  prices  is  that  they  continued  to 
copy  fairly  faithfully  the  course  of  profits,  to  the  point  of  disregarding  the 
lowering  of  margin  requirements  that  came  into  effect  on  Nov.  1,  1937, 
and  the  relaxation  of  the  rules  about  margin  trading  that  was  decreased 
in  December, 

1  Then  they  fell  off  to  75  in  the  third  and  45  in  the  fourth  quarter.  The  figure  for  the 
first  quarter  of  1938  is  87  millions  and  for  the  second,  68  millions  a  month.  But  issues  for 
refunding  bank  debts  are  included. 


1018  BUSINESS  CYCLES 

If  wholesale  prices  (B.L.S.  index  of  all  commodities)  may  be  said, 
according  to  the  guess  we  ventured  to  make  in  an  earlier  place,  to  have 
in  1933  made  up  for  as  much  of  their  fall  as  was  due  to  the  preceding 
spiral,  they  should  then  have  resumed  their  downward  course.  They 
would,  in  fact,  have  done  so  but  for  the  rise  that  was  forced  upon  farm 
products  and  foods:  the  index  of  the  prices  of  other  commodities  declined 
until  the  second  quarter  of  1935.1  The  slow  and  hesitating  increase 
in  the  latter,  which  then  began  and  after  a  setback  continued  until 
October  1936,  was  exactly  what  we  should  have  expected  for  the  pros- 
perity phase  of  a  Juglar  located  at  this  one  is  within  the  Kondratieff: 
we  know  why  in  this  particular  pattern — compare  the  eighties — prices 
offer  a  strong  resistance  to  any  forces  that  strive  to  raise  them,  to  the 
internal  ones  which  originate  in  the  mechanism  of  prosperity,  as  well  as  to 
any  external  ones,  political  or  other. 

But  precisely  because  of  this  we  cannot  agree  with  those  observers 
who  hailed  the  violent  rise  which  occurred  in  the  fourth  quarter  of  1936, 
as  the  sign  that  prosperity — let  alone  "recovery" — had  come  at  last. 
From  the  standpoint  of  our  process  that  was,  on  the  contrary,  abnormal 
and  calls  for  external  explanation,  which,  of  course,  is  not  far  to  seek. 
That  rise  of  prices  heralded  not  prosperity  but**  inflation."2  The  gears 

1  Textiles  and  hides  and  leather  in  1935  reacted  by  a  rise  to  their  fall  in  1934,  the  former 
not  quite  reaching,  the  latter  surpassing  the  level  obtaining  at  the  beginning  of  1934. 
Metals  and  metal  products,  building  materials,  house  furnishings  hardly  changed  at  all 
until  the  general  rise  in  the  last  quarter  of  1937,  and  other  groups,  though  displaying  more 
fluctuations,  behaved  similarly. 

We  shall  in  the  text  confine  discussion  to  the  behavior  of  the  Other  Commodity  Index, 
because  agrarian  prices  were  too  much  influenced  by  public  policy  to  be  relevant  to  our 
discussion.  It  could  be  shown,  however — perhaps  it  is  even  obvious — that  they  did  not 
entirely  fail  to  reflect  our  process.  We  will  recall  that  farm  products  followed  up  their 
spectacular  rise  in  1933  by  an  almost  equally  strong  one  in  1934  and  the  first  quarter  of 
1935.  From  May  1935  to  May  1936  they,  on  the  whole,  declined;  but  cash  receipts  from 
sales  for  1935  were,  nevertheless,  well  above  those  for  1934.  From  June  1936  to  March 
1937  we  have  another  spectacular  rise  in  the  index  of  farm  products,  resulting  in  a  still 
larger  cash  revenue  from  sales,  and  for  1937  the  latter  again  increased  by  about  7  per  cent, 
in  spite  of  the  precipitous  fall  which,  from  April  1937  to  March  1938,  carried  that  index  to 
below  the  level  of  the  last  quarter  of  1934.  The  continuation  of  the  back-to-the-farm 
movement,  which  was  revealed  by  a  special  survey  of  the  division  of  agriculture  of  the 
census  bureau  (in  cooperation  with  the  bureau  of  agricultural  economics,  1938)  is,  hence, 
not  surprising,  though  interpretation  must  differ  from  that  for  the  same  movement  during 
1930  to  1932. 

2  The  meaning  of  that  term  is  obvious  in  this  case.     And,  however  we  may  dislike  the 
word,  there  is  no  other  which  in  this  connection  would  be  equally  suggestive  of  a  self- 
reinforcing  sequence  of  increases  in  monetary  values,  receipts,  and  costs  that  will,  unless 
stopped  from  outside,  go  on  indefinitely,  each  step  enforcing  the  next  and  defeating  the 
effects  of  the  preceding  one.     The  role  of  armament  demand  in  the  rise  of  some  prices 
(copper,  scrap,  and  so  on)  as  well  as  in  other  features  of  that  prosperity  (operating  rate  of 
the  steel  industry)  must  not  be  forgotten,  however. 


THE  WORLD  CRISIS  AND  AFTER  1019 

of  the  engine  composed  of  public  spending  and  newly  created  facilities 
for  credit  expansion  began  to  mesh.  The  case  illustrates,  as  regards  the 
former,  the  proposition  that  income-generating  expenditure  may  raise 
the  price  level  in  the  presence  of  underutilized  resources  and,  as  regards 
the  latter,  the  proposition  that  increasing  the  lending  power  of  the  bank- 
ing system  does  next  to  nothing  in  the  depression  which  it  is  intended  to 
remedy,  and  very  little  in  recovery,  but  takes  effect  when  it  is  not  intended 
to  do  so,  viz.,  in  prosperity.  It  equally  illustrates  the  mechanism  that 
works  by  two  levers,  anticipations  on  the  one  hand  and,  what  is  more 
important,  a  race  between  prices  and  wages1  on  the  other.  Of  course, 
only  fractional  use  was  actually  made  of  the  powers  of  the  deposit-manu- 
facturing machine,  but  this  is  no  objection  to  that  diagnosis.  And  it  is 
significant  to  observe  how  much  of  it  went  into  financing  of  households* 
expenditure.  Banks,  trying  to  find  outlets  for  their  idle  cash  and 
responding  to  the  incessant  appeals,  underlined  by  threats,  that  they 
should  lend  more  freely,  sometimes  went  to  the  length  of  inviting  applica- 
tions for  personal  loans  by  newspaper  advertisements,  but  more  com- 
monly financed  intermediate  lending  agencies  and  retailers'  receivables, 
from  installment  paper  to  open  accounts.  Cash  lending  by  other  types  of 
lenders  also  increased.  Thus,  consumers*  credit  which  had  been  reduced 
by  depression  rose  again  to  the  1930  figure  or  nearly  so.  We  are  not  going 
to  reopen  the  theoretical  problem  of  consumers*  credit.  It  is  enough  to 
point  to  the  fact,  its  relation  to  prices,  and  its  importance  in  any  relapse, 
which  it  may  be  sufficient,  long  before  any  great  percentage  of  households 
default,  to  turn  into  a  vicious  spiral. 

The  powers  of  credit  creation  being  what  they  were,  that  process 
could  not  only  have  gone  on  indefinitely  but  also  at  a  pace  beyond  the 
possibilities  of  expanding  physical  output.  We  cannot,  therefore, 
completely  rely  on  the  automatism  of  our  process  for  full  explanation  of 
the  precipitous  fall  of  prices  in  the  last  quarter  (the  rise  ended  with  the 
first)  that  rapidly  tapered  off  in  the  first  half  of  1938.  But  expectation 
from  our  model  is  for  a  fall  during  recession  phases,  and  the  recession  of 
a  Juglar  that  runs  its  course  on  a  Kondratieff  downgrade  should  end  up 
with  a  price  level  lower  than  that  of  the  neighborhood  from  which  the 
Juglar  rose.  If  the  fall  that  actually  occurred  was  from  a  higher  level 
and  hence  steeper  than  can  be  explained  by  our  process,  the  eventual 
result  was  not  so  different  from  what  it  should  have  been  if  nothing  else 
had  contributed  to  it.  We  should  even  have  expected  a  decline  to  below 

1  If  such  a  race  is  not  allowed  to  start,  results  may  be  different  and  more  akin  to  those 
predicated  by  the  better  type  of  "reflation"  theories.  The  German  and  Japanese  cases 
are  conspicuous  and  the  English  case  is  a  less  conspicuous  instance  of  the  success  that  may 
attend  a  corresponding  policy  under  appropriate  conditions.  For  further  comment  see 
below  sub  5, 


1020  BUSINESS  CYCLES 

the  figure  of  the  second  quarter  of  1935.  No  group  save  farm  products, 
foods,  and  textiles  fell  as  much  as  that,  and  the  index  of  commodities 
other  than  farm  products  and  foods  never  returned  even  to  its  annual 
figure  for  1935.  There  is  a  strong  presumption  that,  barring  monetary 
management  and  other  price-raising  policies — policies,  that  is,  which  are 
price  raising  in  effect,  whatever  the  intentions  and  phraseologies — price 
level  would  continue  to  fall,  though  gently,  for  more  than  another 
decade. 

It  may  be  noted  that  of  the  784  commodities  which  enter  into  the 
B.L.S.  all-commodity  index,  the  189  items  that  fell  most  from  1926  to 
1933  (when  their  prices  ranged  from  5.4  to  42.9  per  cent  of  the  1926 
figures)  were  at  the  end  of  1937  on  the  average  at  about  60  per  cent.  Of 
these  all  but  10  belong  in  the  categories  of  extractive  raw  material, 
agricultural  products,  and  little-processed  staples.  The  190  which  fell 
least  (and  in  1933  stood  at  between  78.1  and  118.3  per  cent  of  their  1926 
prices)  were  at  the  end  of  1937  on  the  average  at  about  100  per  cent. 
Highly  finished  articles  form  the  bulk  of  this  group,  such  as  chemicals, 
agricultural  implements,  and  so  on.  The  implications  of  this  are  in  many 
cases  weakened,  in  some  reversed,  by  taking  account  of  changes  in  quality. 
In  others,  special  conditions  explain  the  "rigidity."  In  no  case  is  it  possi- 
ble to  infer  from  these  facts  alone  anything  about  lack  of  balance  in  the 
price  system  or  about  lack  of  flexibility  in  prices  per  service  unit.1  The 
relative  fall  in  raw-material  prices  foreshadows,  and  is  a  condition  of,  a 
new  equilibrium  at  vastly  increased  figures  of  output. 

3.  There  is  thus  some  justification  for  going  on  to  speak  of  Juglar 
phases.  Our  main  reason  for  doing  so  lies,  however,  in  the  nature  of  the 
industrial  processes  of  the  period.  This  becomes  apparent  if  we  ask  our- 
selves what  we  should  have  expected  to  happen.  We  may  think,  for 
example,  of  our  experience  with  the  railroad  Juglars  of  the  nineteenth 
century.  They  all  had  a  family  likeness  and  were  quantitatively  dom- 
inated by  railroad  construction  not  only  as  long  as  this  was  the  funda- 
mentally new  thing  but  also  for  a  time  of  completing  development. 
Historical  record  does  not  lead  us  to  expect  that  innovations  of  the  first 
magnitude — in  a  financial  sense — will  turn  up  in  Kondratieff  downgrades : 
every  railroad  Juglar  had,  besides  marking  a  step  in  the  evolution  of  the 
railroad  system,  its  own  contribution  of  novelties  to  make;  but  quan- 
titatively they  were  of  minor  immediate  importance — even  Bessemer 
steel  was,  for  instance — as  compared  with  the  innovation  that  made  the 
railroad  Kondratieff.  We  know  what  made  the  current  one  and  are 
thus  in  a  position  to  form  a  definite  "forecast"  in  order  to  compare  it 
with  the  actual  course  of  things.  We  also  know  that  downgrades  are 

1  On  this  subject  see  Professor  E.  S.  Mason's  excellent  article  on  Price  Inflexibility, 
Review  of  Economic  Statistics,  May  1988, 


THE  WORLD  CRISIS  AND  AFTER  1021 

characterized  by  very  numerous  small  and  induced  innovations.  To 
these  it  is  impossible  to  do  justice.  But  the  great  lines  are  simple  enough 
to  list. 

In  doing  so  we  do  not  on  the  whole  meet  with  disappointment  at  the 
first  step.  This  is,  first  and  foremost,  the  Kondratieff  of  electricity. 
The  current  Juglar  should  have  carried  on  the  work  of  its  predecessors 
at  least  as  much  as  the  fifth  Juglar  of  the  second  Kondratieff  carried 
on  the  railroad  work  of  the  four  that  went  before  it,  at  least  because 
investment  opportunity  seems  even  greater  in  this  case,  considering  the 
work  to  be  done  within  the  range  of  present  technical  and  economic 
vision.1  The  production  of,  and  the  innovations  in,  electrotechnical 
manufacturing,  in  fact,  come  fully  up  to  expectation,  so  obviously  that 
we  need  not  stay  to  prove  it.  Kilowatt-hours  produced  passed  the  1929 
mark  in  1935.  Power  developments  have  consisted  chiefly  in  progress 
with  the  great  public  ventures:  Boulder  Dam,  Bonneville,  Grand  Coulee, 
Fort  Peck,  and  Muscle  Shoals,  which  are  to  increase  capacity  installed 
by  over  4  billion  kilowatts,  and  in  the  smaller  projects  sponsored  under 
the  Rural  Electrification  Act  of  1936,  mainly  by  rural  cooperatives. 
But  the  amount  of  construction  done  by  privately  owned  public  utilities 
is  indeed  disappointing.  We  should  have  expected  vigorous  expansion  in 
power  plants,  substations,  and  transmission  lines,  and  the  pecuniary 
investment  corresponding  to  this  and  to  the  incident  expansion  of 
equipment  should  have  contributed  decisively  to  the  processes  of  pros- 
perity. Whether  or  not  the  failure  of  the  actual  development  (which 
was,  however,  not  negligible)  to  bear  out  those  expectations  is  to  be 
recorded  against  the  principles  that  yield  them  depends  on  whether  or  not 
it  is  adequately  accounted  for  by  some  inhibiting  factor  external  to  the 
industrial  organism.  This  will  be  touched  upon  later. 

In  the  second  place,  this  is  the  Kondratieff  of  the  automobile.  No 
such  development  as  that  of  the  motorcar  industry  has  ever  broken  off 
suddenly.  Therefore  the  current  Juglar  should  include,  or  partly  consist 
in,  another  automobile  wave.  As  everyone  knows,  we  are  again  not 
disappointed  in  this  expectation.  Such  prosperity  as  there  was  clearly 
centered  in  the  motor  industry  and  its  satellites,  such  as  tires  and  inner 
tubes,  plate  glass,  steel,  by-product  coke,  and  gasoline,  and  this  accounts 
for  the  characteristic  inequality  of  the  upswing  as  between  industries 
which  was  evident  even  in  1936  when  improvement  had  become  general. 
Automobiles  (cars  and  trucks;  United  States  and  Canada)  recrossed  the 
4  million  unit  line  in  1935  and  the  5  million  unit  line  in  1937,  when  produc- 

1  In  1985  only  800,000  of  the  6,800,000  farms  were  supplied  with  electricity.  Urban 
domestic  consumption  was  678  kilowatt-hours  per  home,  whereas  according  to  Mr.  Samuel 
Ferguson  it  might  be  8,400  kilowatt-hours  (?).  It  increased  by  50  per  cent  between  1920 
and  1936. 


1022  BUSINESS  CYCLES 

tion  was  higher  than  in  any  previous  year  excepting  1929,  or,  if  we  take 
account  of  the  decline  in  the  number  produced  for  export  and  merely 
consider  production  for  domestic  consumption,  about  94  per  cent  of  the 
1929  figure.  It  is  true  that  owing  to  the  slump  in  the  second  half  of  the 
year  part  of  this  output  merely  went  into  dealers'  stocks,  but  on  the  other 
hand,  it  must  be  remembered  that  part  of  the  increase  in  the  motorization 
of  the  country  occurred  in  rural  communities  (82  per  cent  of  farmers 
owning  cars  according  to  the  Bureau  of  Home  Economics  of  the  Depart- 
ment of  Agriculture),  the  demand  from  which  was  to  two-thirds  satisfied 
by  second-hand  cars.  The  essential  point,  however,  is  that  the  industry 
was  not  simply,  as  we  have  expressed  it,  "drawn  along**  by  environ- 
mental growth  or  improvement  (roads)  and  that  it  not  merely  grew  into 
existing,  but  also  created  new  economic  space.  The  changes  in  the 
product  were  not  merely  routine  changes  in  design  and  so  on,  such  as 
occur  in  operating  any  textile  mill,  but  included  also  a  number  of,  if 
individually  minor,  innovations — 33,721  patents  were  issued  in  this  field 
since  1934 — such  as  the  turret  top  (all-steel  frame),  automobile  radios, 
knee-action  suspension,  nonfading  finishes,  and  others,  among  them  still 
"incubating'*  ones  as,  for  instance,  the  pancake  motor. 

Moreover,  the  industry  or  some  of  its  concerns  reached  out  into  neigh- 
boring fields  and  became  responsible  for  innovations  in  these — the  Diesel- 
engine  division  of  General  Motors  (new  plants  in  La  Grange  and  Detroit), 
which  had  much  to  do  with  the  increase  in  speed  of  trains  during  the 
period,  and  the  same  concern's  activities  in  the  fields  of  refrigeration, 
air  conditioning,  small-scale  light  arid  power  production,  and  aviation 
(Delco  Frigidaire  and  Conditioning  Division,  Delco  Radio  Division,  Delco 
Electric  Light  and  Power  Plants,  Bendix  Aviation  Corporation,  North 
American  Aviation)  may  be  cited  as  examples.  Considerable  outlay  for 
current  developments,  as  well  as  for  new  and  improved  plant  capacity, 
attended  this  development.  The  "  competing-down  "  process  is  evi- 
denced by  the  increasing  share  in  production  of  the  three  leading  concerns. 
Behavior  of  stock  prices,  wages,  and  prices  of  product  conform  to  our 
general  idea  of  an  industry  that  was  still  innovating  and  expanding  under 
the  impulse  of  innovations.  Average  hourly  rate  excluding  increase  in 
payment  for  overtime  increased,  for  example,  by  nearly  20  per  cent  in 
1937  and  was  then  over  28  per  cent  above  the  1929  level.1  As  to  price, 
there  is  the  usual  difficulty  about  quality.2  Retail  sales  value  in  1937 
was  roughly  3.85  billions  as  compared  with  about  4.77  billions  in  1929. 

The  allied  industries  all  display,  though  to  a  varying  degree,  the 
same  characteristics.  For  the  rubber  industry,  in  particular,  every  one 

1  In  the  first  half  of  1938  the  average  rate  was  a  little  over  90  cents  an  hour.  The 
working  week  was  (April,  1938)  a  little  over  31  hours. 

8  Discounts  and  trading-in  allowances,  moreover,  make  quoted  prices  all  but  meaningless. 


THE  WORLD  CRISIS  AND  AFTER  1028 

of  the  above  statements  could  be  paralleled.1  Innovation  was  of  a 
similar  type  ("  supertwist "  again,  the  tractor  and  implement  tire,  output 
of  which  in  1936  was  1,775  per  cent  above  that  of  1933,  activities  in 
rubber  and  even  in  cotton  growing  and  cotton  milling),  but  price  and 
effect  on  output  must  be  corrected  for  the  increased  amount  of  service 
units  contained  in  a  modern  tire  of  good  quality — according  to  an  esti- 
mate by  the  industry  on  an  average  31,446  miles  in  1937  as  compared 
with  18,546  in  1929,  which  would  make  a  fall  in  price  per  mile  of  some- 
thing like  38  per  cent.  Wage  rates  also  make  a  similar  showing. 

Improved  quality  and  more  economical  use  progressively  deprive 
steel  output  of  its  value  as  a  cyclical  thermometer.2  Only  if  this  is  taken 
into  account,  do  the  33.4  million  tons  of  ingots  produced  in  1935  or  the  46.9 
millions  produced  in  19363  acquire  their  true  significance  and  some  com- 
parability with  the  figures  of  the  preceding  Juglar  prosperity,  and  only 
then  is  it  possible  to  recognize  the  output  of  the  first  five  months  of  1937 
as  truly  indicative  of  Juglar  high  tide.4  The  relative  importance  of  the 
lighter  steel  products,  steel  sheets,  strip  wire,  tin  plate,  and  so  on  which, 
though  they  enter  into  farm  implements  and  machinery  in  general,  may 
be  said  to  be  more  nearly  consumers'  goods'  material  was,  of  course, 
greater  than  at  any  previous  time — absolute  output  was  also  at  record 
figures  in  1936  and  during  the  first  three  quarters  of  1937 — but  its  varia- 
tions yet  indicate  the  course  of  cyclical  subphases:  lighter  products  were 
47  per  cent  of  the  total  in  1935,  only  42.5  per  cent  in  1936  when  prosperity 
had  got  into  its  stride,  and  about  43  per  cent  in  the  first  half  of  1937 
when  it  was  tapering  off.  The  new  advance  in  rolling  ("continuous 
mills"),  foreshadowed  at  the  threshold  of  the  depression,  constituted 
the  most  conspicuous  innovation,  but  there  was  a  large  number  of  smaller 

1  According  to  the  Automobile  Manufacturers'  Association's  estimate,  sales  value  of 
automobile-tire  shipments  (including  casings,  inner  tubes,  solids,  cushions,  and  sundries) 
was  298  millions  in  1932,  307  millions  in  1933,  324  in  1934,  384  in  1935,  436  in  1936,  and 
thus  in  1936  still  below  not  only  1929  (722)  but  also  1930  (532). 

2  But  against  the  facts  that  a  pound  of  steel,  as  we  have  put  it  elsewhere,  goes  in  general 
much  further  than  it  used  to  even  in  the  twenties,  and  that  the  share  of  high-quality  special 
steels  and  steel  alloys  increases,  must  be  set  the  other  fact  that  an  ingot  yields  a  smaller 
amount  of  finished  product  in  the  case  of  this  high-quality  steel.     Iron  ore  consumed  is, 
of  course,  influenced  by  the  rapid  increase  in  the  use  of  scrap. 

3  Capacity  per  head  of  population  was  then  at  the  lowest  figure  since  1929,  but  for  the 
reason  mentioned  this  also  does  not  mean  what  it  seems  to.     Its  maximum  occured  in 
1934,  after  which  it  was  reduced  by  wholesale  scrapping,  another  indication  of  rapid 
technological  advance. 

4  But  even  apart  from  those  considerations,  ingot  output  of  January  1937,  when  the 
industry  operated  at  86  per  cent  of  capacity,  was  nearly  at  record  high,  since  it  was  only 
exceeded  by  the  output  during  a  few  months  of  1929,  which  was  clearly  exceptional. 
Moreover,  it  should  be  remembered  that  also  after  1873  peaks  of  pig-iron  production 
remained  below  that  for  1872  for  quite  a  while. 


1024  BUSINESS  CYCLES 

ones,  principally  in  the  field  of  alloys  and  other  specialties  (flat-rolled 
steel),  but  also  in  others,  some  of  which,  such  as  the  progress  in  welding, 
were  effective  in  creating  new  markets.  The  amounts  reported  as  spent 
by  the  industry  on  new  construction  and  equipment,  a  little  less  than  700 
millions  for  1935  to  1937,  do  not  indicate  more  than  that  there  was  non- 
negligible  investment,  for  that  figure,  on  the  one  hand,  includes  some 
replacements  and  mere  extensions  while,  on  the  other  hand,  additional 
allowance  should  be  made  for  improvements  financed  under  other  head- 
ings. New  blast  furnaces — three  in  1937,  two  of  them  replacing,  though 
of  course  with  improvements,  dismantled  ones — coke  ovens  for  a  million 
tons  beyond  replacement,  open-hearth  furnaces,  electric  furnaces,  and 
then  the  new  rolling  mills  mentioned  constitute  the  main  items.  Behav- 
ior of  prices,  employment,  and  wages  also  conforms  to  expectation. 
Composite  price  of  steel  (American  Iron  and  Steel  Institute),  which  was 
$67.71  per  ton  for  1923  and  $47.41  for  1933,  increased  to  $56.85  in  1937, 
average  hourly  rates  were  respectively  59.6,  52.4,  and  (from  Mar.  16, 
1937)  about  83  cents,  well  above  the  industrial  average  and  above  1929. 
Employment  (number  of  wage  earners;  not  man-hours)  increased  in 
spite  of  the  labor-saving  nature  of  some  of  the  innovations,  and  was  about 
30  per  cent  above  the  1929  level  in  the  summer  of  1937  (over  three  times 
the  number  in  1879,  which  is,  however,  a  census  figure  and  not  entirely 
comparable) . 

In  appraising  steel  developments  and  prospects,  armament  and 
construction  demand  must  be  taken  into  account. *  The  latter  comes  in 
as  a  negative  item,  i.e.,  the  cyclical  significance  of  steel  production  must 
be  interpreted  in  the  light  of  the  fact  that  this  component  failed  to 
contribute  as  much  as  we  should  have  expected.  We  have  noticed  the 
fact  and  one  of  its  causes,  but  an  additional  remark  suggests  itself.  We 
would  not  per  se  have  expected  a  particularly  high  wave  of  residential 
building  during  a  prosperity  phase,  although  this  is  more  likely  to  occur — 
as  it  did  occur  in  1925  and  1926 — in  the  downgrade  than  in  the  upgrade 
of  a  Kondratieff.  But  we  miss  enterprise  which  in  this  case  there  were 
particular  reasons  to  expect.  The  mass  production  of  the  cheap, 
prefabricated  house  is  one  of  the  most  obvious  innovations  of  the  present 
and  near  future,  and  it  should  if  anything  have  been  promoted  by  high 

1  It  is,  presumably,  also  necessary  to  take  account  of  the  fact  that  the  demand  for 
railroad  equipment,  which  materially  contributed  in  1984  and  after,  was  greater  than  was 
warranted  by  the  results  and  prospects  of  the  railroad  business,  which  continued  throughout 
recovery  and  prosperity  to  illustrate  our  competing-down  process.  Operating  revenue 
from  freight  of  Class  I  railroads,  which  at  its  maximum  of  1929  was  only  10  per  cent  above 
1920,  fell  to  below  60  per  cent  in  1932  (minimum)  and  was  not  quite  75  per  cent  at  the  end 
of  1936.  Revenue  from  transportation  of  passengers  fell  all  along  except  in  1923,  reached 
80  per  cent  of  1920  in  1932,  was  still  lower  in  1933  and  only  very  little  over  30  per  cent  at 
the  end  of  1936  (below  30  per  cent  for  the  year). 


THE  WORLD  CRISIS  AND  AFTER  1025 

building  costs.  Such  enterprises  first  emerged  at  the  threshold  of  the 
current  Kondratieff  (City  and  Suburban  Homes,  New  York,  1896, 
Washington  Sanitary  Improvement  Company,  at  the  same  time)  and 
prefabrication,  domestic  electrification,  steel  developments,  and  other 
innovations1  are  so  many  propelling  factors.  Yet  nothing  of  that  kind 
happened  on  any  scale,  at  least  until  the  passage  of  the  Federal 
Housing  Act  in  1936  setting  up  a  Federal  Housing  Administration 
which  was  authorized  to  "insure"  mortgages  for  projects  up  to  10 
millions. 

The  chemical  industry  does  not  disappoint  us.  Expansion,  innova- 
tion, and  investment  were  on  a  considerable  scale — advance  in  the  fields 
of  "synthetic  organics,"  of  refrigerants,  of  protective  coatings  (which 
also  gave  an  impulse  to  the  paper  industry),  or  of  plastics  may  indicate 
a  type  of  improvements  which  sum  up  to  a  very  substantial  item — both 
within  and  without  the  two  big  concerns.  The  rayon  industry  out- 
stripped previous  records  partly  by  conquering  new  uses  or  markets  and 
partly  by  technological  progress,  which  included  a  major  innovation. 
The  standard  fiber  did  not  even  begin  to  show  its  possibilities — and 
threats.  New  in  our  sense  was  air  conditioning.  Although  installations 
started  in  1919,  they  reached  a  maximum  value  of  but  17  millions  in 
1930.  The  industry  was  still  in  the  experimental  stage.  It  was  only  in 
the  prosperity  of  the  current  Juglar  that  it  rose  to  the  modest  heights  of 
35  millions  in  1935,  of  53  in  1936,  and  about  85  in  1937.2  Aviation  may 
be  said  to  have  reached  about  that  stage  in  which  railroads  were  in  the 
thirties  of  the  nineteenth  century.  Aircraft  making  enjoyed  a  great 
boom  every  year,  1938  also  included,  setting  a  new  record,  mainly, 
however,  because  of  military  demand,  which  accounted  for  about  60 
per  cent  of  1937  sales  and  for  most  of  the  highly  profitable  exports  which 
amounted  to  34  per  cent  of  1937  sales.  Even  so,  the  decisive  tech- 
nological advance  achieved  hardly  began  to  unfold  its  effects,  and  total 
sales  (109  millions  in  1937)  and  profits  of  the  eight  major  concerns  were 
distinctly  modest.  Still  more  so  was  the  progress,  in  everything  but 
equipment  and  quality  of  service,  of  the  three  great  operating  companies, 
whose  operating  revenue  for  1937  did  not  exceed  25  millions,  spelling 
deficit  in  all  three  cases.  Two-thirds  of  their  total  investments  of  about 
120  millions  would  ex  visu  of  that  year  have  to  be  considered  as  lost. 
Many  reasons  besides  the  peculiar  difficulties  incident  to  this  industry 
(and  the  series  of  disasters)  account  for  this.  But  since  its  lack  of 

1One  is  even  fundamentally  new,  the  "cotton  house."  Mechanization  of  the  actual 
building  was  also  further  developed,  but  there  was  no  new  departure  there.  Improvement 
in  assembling  parts  still  lags  behind  improvement  in  the  parts  themselves. 

2  (Incomplete)  estimates  from  figures  of  principal  firms  by  LaRue  Applegate,  The 
Annalist,  Feb.  12,  1937,  p.  268. 


1026  BUSINESS  CYCLES 

quantitative  significance  for  the  current  Juglar  is  obvious,  we  need  not 
touch  upon  its  problems. 

4.  Having  thus  satisfied  ourselves  that  the  processes  which  in  the 
past  used  to  carry  prosperities  have  not  been  absent  in  the  present 
instance,  we  have  established  a  right  to  speak  of  a  Juglar  prosperity  and 
to  infer  from  experience  that  it  would  have  asserted  itself  without  any 
external  impulse  being  imparted  to  the  system  by  government  expendi- 
ture or  any  other  factor.  In  particular,  there  is  nothing  %to  indicate  that 
objective  opportunities  were  smaller  or  capitalist  motivation  weaker 
than  they  had  been,  say,  in  1925.  The  problem  why  that  prosperity 
was  so  weak,  and  why  it  should  have  been  followed  by  so  severe  a  slump 
now  emerges  in  its  proper  setting. 

At  the  outset  we  dismiss  the  possibility — which,  in  fact,  has  not  been 
sponsored,  so  far  as  the  writer  knows,  by  any  economist — that  the  steps 
taken  toward  freer  trade  (Montevideo  conference  of  American  states  and, 
partially  in  fulfilment  of  the  pledge  there  made,  the  Reciprocal  Trade 
Agreements  Act  of  June  12,  1934,  and  the  Presidential  Proclamation  of 
July  8,  1935)  can  have  materially  dampened  prosperity  or  intensified 
depression.  Of  the  17  agreements  actually  entered  into  up  to  March 
1938,  only  those  with  Cuba  (September  1934)  and  Canada  (January 
1936)  can  possibly  have  exerted  nonnegligible  effects  in  spots,1  and  from 
what  depressive  effects  they  did  exert  it  would  be  necessary  to  deduct 
gains  in  other  sectors.  There  may  even  have  been  a  net  contribution  to 
recovery  and  prosperity.  If  so,  it  cannot  have  been  significant,  however. 
Exports,  which  in  1929  had  amounted  to  5,241  millions,  rose  from  the 
1932  figure  of  1,611  only  to — in  millions  of  full-weight  (or  devaluated) 
dollars— 1,280  (2,133)  in  1934  and  to  1,370  (2,283),  1,474  (2,456),  and 
1,977  (3,295)  in  the  subsequent  years.  And  the  effects  of  general  improve- 
ment in  the  world,  of  devaluation,  and  of  armament  demand  account 
for  the  bulk  of  that. 

It  is  less  obvious  that  the  momentous  changes  in  the  sphere  of  money 
and  credit,  which  have  been  the  subject  of  so  much  controversy,  did  not, 
except  by  facilitating  government  expenditure,  decisively  influence 
economic  processes  during  the  years  under  survey.  In  order  to  establish 
this,  we  need  not  go  into  the  principles  involved  in  those  changes  or  into 
the  question  of  what  long-run  results  are  likely  to  be  in  the  future.  We 
have  indeed  noticed  the  effects  on  money  rates,  on  price  level,  and  on 
consumers'  credit.  But  the  lessons  administered  by  an  experience 

1  Even  effect  "in  spots"  is  in  some  cases  doubtful.  American  shoe  producers  were 
probably  no  unbiased  judges  of  the  dangers  that  threatened  from  the  Bata  concern  (at 
least,  inasmuch  as  it  remained  outside  the  United  States)  or  steel  producers  of  the  danger 
to  their  lives  from  an  "increase  in  imports  of  173  per  cent"  in  the  first  year  of  the  Belgian 
agreement,  which  turns  to  have  been  1.5  millions. 


THE  WORLD  CRISIS  AND  AFTER  1027 

extending  over  a  century  and  a  half  and  the  scarcity  of  acceptable 
applications  for  credit — admitted,  in  the  end,  even  by  fact-finding 
agencies  set  up  for  the  purpose  of  convicting  banks  of  restiveness — pre- 
vented excesses  which  in  fact  were  not  at  the  time  encouraged  by  the 
federal  reserve  authorities.1  The  Banking  Act  of  Aug.  23,  1935,  codified 
and  made  permanent  the  chief  innovations  previously  introduced,  but 
in  doing  so  it  emphasized  the  restrictive  rather  than  the  expansive 
element  in  them.  Following  upon  the  removal  of  the  restrictions 
previously  imposed  upon  transactions  in  foreign  exchange  (Nov.  12,  1934) 
it  was  widely  understood  in  that  sense,  in  spite  of  the  criticism  it  met  from 
"sound-money"  quarters.2  Later  on,  it  was  proved  that — at  least  for 
once — the  brakes  of  the  engine  were  not  mere  window  dressing.  The 
only  problem  relevant  to  our  subject,  therefore,  concerns  the  use  which 
was  made  of  them  and  which  many  people  held  responsible  for  the  occur- 
rence or  the  severity  of  the  slump. 

Faced  with  an  influx  of  gold,  as  persistent  as  it  was  natural — since, 
barring  all  extraeconomic  factors,  which  of  course  helped,  any  commodity 
will  go  to  where  it  is  being  overpaid — and  perhaps  somewhat  concerned 
about  the  violent  rise  in  prices,  the  Treasury  and  the  Federal  Reserve 
Board  took  action.  The  Treasury,  not  prepared  to  go  back  upon  devalu- 
ation, entered  at  the  time  of  the  French  devaluation  (Sept.  28,  1936)  into 
the  Tripartite  Agreement  with  England  and  France,  undisclosed  oper- 
ations under  which  helped  to  control  the  gold  movement  temporarily. 
Moreover,  in  December  1936,  it  inaugurated  the  gold  sterilization  plan:  by 
directly  acquiring  and  impounding  newly  received  gold  it  prevented, 

1  See,  for  example,  the  "conservative"  reply  of  the  board  of  governors  of  the  Federal 
Reserve  System  to  the  chairman  of  the  Senate  Committee  on  Agriculture  and  Forestry, 
published  in  the  Federal  Reserve  Bulletin  of  September  1987. 

2  The  act  did  allow  the  small  state  banks  outside  the  system  to  enjoy  insurance  of 
deposits  without  joining.     It  did  nothing  about  the  division  of  responsibility  for  monetary 
policy  between  the  Treasury  and  the  Board.     It  increased  rather  than  diminished  the 
Board's  independence,  such  as  it  was.     It  developed  the  tool  of  open-market  operations, 
sanctioned  the  Board's  power  to  vary  reserve  requirements,  and  otherwise  strengthened  its 
hold  on  the  policy  of  member  banks.     It  failed  to  include  the  clause  in  the  House  bill  about 
"stabilizing  business."     It  retained  the  gold  cover  requirements  for  Federal  reserve  notes 
and  deposits.     It  also  retained,  although  it  made  them  illusory,  eligibility  requirements. 
It  no  doubt  encouraged  long-term  lending  by  member  banks,  but  not  more  than  it  had 
been  encouraged  before.     It  perpetuated  deposit  insurance.     Much  irrelevant  discussion 
arose  from  confusions  between  the  technically  unsatisfactory  working  of  eligibility  require- 
ments and  the  principle  which  they  faultily  expressed;  between  responsible  practice  of 
member  banks  and  laissez  faire;  and  between  their  functions  in  the  economic  process  and 
such  things  as  their  "duties  to  depositors."     Fundamental  truth  was  often  clothed  in 
wrong  and  fundamental  error  in  correct  arguments.     Solicitude  for  public  morals  and 
public  welfare  often  vested  the  garb  of  group  interests,  antisocial  tendencies  and  levity 
the  garb  of  "  progress! veness."     On  the  whole,  however,  the  practice  envisaged  by  the  act 
as  it  stands  is  still  serious  banking. 


1028  BUSINESS  CYCLES 

up  to  September  1937,  any  effects  of  the  influx  on  bank  reserves  and 
deposits,  the  increase  in  monetary  gold  stock  and  treasury  cash  going 
into  an  Inactive  Gold  Account — an  important,  if  negative,  measure. 

The  Board,  using  its  new  powers,  directly  operated  upon  the  incubus 
of  excess  reserves  by  raising  reserve  requirements,  first  by  50  per  cent 
(effective  Aug.  16,  1936)  and  then  again  by  16%  and  4^f  per  cent 
(effective  Mar.  1  and  May  1,  1937).  Total  and  excess  reserves  had  been 
rising  steadily  from  the  beginning  of  1933  for  all  other,  and  from  the 
beginning  of  1934  also  for  New  York  City  member  banks*  thus  absorbing 
much  of  the  flood  and  once  more  illustrating  the  value  of  the  theory  that 
banks  are  always  "loaned  up."  Great  as  was  the  gold  influx  in  1934,  it 
was  surpassed  in  1935  by  more  than  50  per  cent  (total  net  gold  imports 
roughly  1.75  billions,  over  half  of  them  from  France).  Together  with 
various  smaller  items  which  need  not  concern  us,  this  spelled  an  increase  of 
nearly  1.9  billions  in  the  monetary  gold  stock — an  increase  of  about 
350  millions  in  money  "in  circulation"  and  an  increase  of  about  1,5 
billions  in  members*  reserves,  or  of  about  1  billion  in  their  excess  reserves. 
Influx  continued,  though  with  abated  force  and  significant  intermissions, 
throughout  1936:  the  monetary  gold  stock  rose  by  more  than  1  billion  and 
members'  reserves  by  almost  as  much,  the  loss  to  circulation  being  almost 
wholly  compensated  by  the  reduction  of  federal  balances  at  reserve 
banks.  In  1937  the  tide  swelled  again,  1.3  billions  producing  an  increase 
in  monetary  gold  stock  of  1.5  billions,  in  spite  of  the  reversal  of  the  move- 
ment in  November  and  December.  But,  as  stated  above,  owing  to  the 
sterilization  policy  this  did  not  go  into  bank  reserves,  which  continued, 
however,  through  1937  to  move  at  the  high  level  attained  toward  the 
end  of  1936,  not  far  from  7  billions  for  all  member  banks.  The  increase 
in  deposits  (adjusted  demand  deposits;  all  reporting  member  banks)  of 
about  2.5  billions  in  1935  and  of  about  1.7  in  1936  substantially  reflects 
government  disbursements  and,  until  May  1936  investments,  thereafter 
other  loans.1  They  began  to  fall  at  the  end  of  the  first  quarter  of  1937 
and  went  on  falling  almost  uninterruptedly,  though  more  in  New  York 
than  Outside,  throughout  that  year.  The  first  part  of  this  development 
until  about  the  middle  of  1936  does  not  call  for  additional  comment;  the 
second  part  brings  us  back  to  our  problem. 

In  our  survey  of  time-series  contours  it  has  been  observed  that  pros- 
perity did  not  bring  about  that  stiffening  of  money  rates  which,  even  in  a 
prosperity  within  a  Kondratieff  depression,  we  should  have  expected. 
The  facts  we  have  just  glanced  at — excess  reserves,  in  particular — amply 
explain  this.  It  has  also  been  pointed  out  that,  given  these  conditions, 
the  microscopic  rise  of  some  rates — commercial  paper  rate  not  among 

1  There  was  a  slight  contraction,  for  the  100  cities  outside  New  York  more  than  slight, 
at  the  beginning  of  1936,  which  was  quickly  made  up  for  and  is  cyclically  understandable. 


THE  WORLD  CRISIS  AND  AFTER  1029 

them — that  occurred  in  1936,  and  the  somewhat  more  perceptible  though 
still  insignificant  increases  that  occurred  in  1937  also  in  bond  yields, 
cannot  primarily  be  explained  by  the  modest  increase  in  business  borrow- 
ing but  was  due  mainly  to  monetary  management.  It  was,  in  fact,  a 
reaction  to  the  increase  in  reserve  requirements — under  conditions  of 
imperfect  competition  curtailment  of  even  an  unsalable  excess  of  supply 
can  have  effect  on  price.1  But  while  monetary  management  produced 
this  effect,  it  certainly  did  not,  via  the  rate  of  interest,  produce  any  other; 
for  no  business  calculation  can  in  practice  be  affected  by  so  minute  an 
increase. 

But  together  with  the  increase  in  all  other  loans,  tne  increase  in  reserve 
requirements  by  50  per  cent  was  the  signal  for  member  banks  to  start 
reducing  their  investments.  The  effectiveness  of  this  signal  can,  how- 
ever, have  been  due  only  to  the  fact  that  member  banks  had  for  other 
reasons  already  begun  to  feel  uneasy  about  their  portfolios.  For  no 
reaction  was  in  itself  enforced  by  that  measure,  which  only  reduced  excess 
reserves  from  about  2.9  billions  (July  15)  to  1.8  (Aug.  19),  because  almost 
simultaneous  treasury  disbursements  absorbed  the  effect  to  the  extent 
of  about  360  millions.2  It  is  true  that  application  of  this  weapon,  as 
shaped,  will  unavoidably  create  some  difficulties  unless  the  reserve  posi- 
tions of  all  members  are  exactly  equal.  But  these  difficulties  were  negli- 
gible in  this  case,  and  very  few  members  had  to  borrow  small  amounts 
from  reserve  banks  at  the  critical  time.  Even  withdrawals  of  funds  from 
New  York  correspondents  did  not  amount  to  much,  nor  did  the  sales  of 
investments  between  the  middle  of  July  and  Aug.  19  (160  millions)  and  in 
November.  Adjusted  demand  deposits  declined  but  slightly  in  August 
and  then  increased  vigorously  again. 

It  is  a  question  of  some  interest  why  the  Board  did  not  leave  things 
at  that.  The  prospective  reduction  in  federal  income-generating 
expenditure  was,  after  all,  no  secret,  and  pro  futuro  the  gold  influx  was 
being  taken  care  of  by  the  sterilization  policy.  The  revival  of  business 
borrowing,  which  kept  up  total  earning  assets  of  members  and  even 
increased  outside  deposits,  and  the  strong  increase  in  money  in  "circula- 
tion" were  no  reasons  for  further  action,  considering  the  phase  of  the 
industrial  process.  Mechanistic  views  about  the  Supply  of  Money  seem 
once  more  to  have  carried  more  than  their  due  weight.  But  even  so,  the 
announcement  on  Jan.  30  of  a  further  increase  in  reserve  requirements  of 
33^<j  per  cent  cannot  be  held  responsible  for  any  depressive  symptoms 
and,  especially,  for  the  rapid  fall  of  corporate  security  issues  in  the  third 

1  This  argument,  of  course,  applies  also  to  the  upward  shift  in  business  demand  for 
loans.     But  it  is  reasonable  to  suppose  that  the  conspicuous  measure  of  raising  require- 
ments was  the  more  potent  factor. 

2  That  is  why  excess  reserves  were  at  about  3.25  billions  on  Aug.  15. 


1030  BUSINESS  CYCLES 

and  fourth  quarters  of  1937.  In  case  the  behavior  of  interest  rates  should 
not  be  considered  sufficient  proof,  we  will  observe  how  well  the  market 
stood  the  experiment.  Some  management  was  indeed  necessary,  especially 
at  the  second  step,  and  transition  was  then  not  quite  so  easy  as  it  had 
been  with  the  50  per  cent  increase.  In  April,  banks  prepared  themselves 
for  that  second  step,  which,  as  it  seemed  at  the  time  of  the  announcement, 
should  have  reduced  excess  reserves  to  500  millions,  but  reduced  them 
only  to  about  850,  owing  to  an  increase  in  total  reserves,  the  decline  in 
deposits,  and  government  disbursements.  The  reserve  system  also  helped 
by  an  open-market  purchase  (96  millions),  and  all  this  raised  excess 
reserves  to  1.6  billions  toward  the  end  of  April.  Thus  prepared,  the 
market  went  through  the  treasury  financing  and  the  quarterly  tax  pay- 
ments of  June  without  any  difficulty.  Not  even  New  York  banks 
(which  one  would  think  should  have  displayed  some  symptoms  of  strain) 
had  to  borrow  from  the  reserve  bank.  Bill  dealers  reduced  rates  on  bank 
acceptances  on  June  22  after  200  millions  of  treasury  bills  had  on  June  17 
and  18  been  repaid  without  replacement.  There  was  some  interbank 
borrowing,  but  to  a  very  small  extent.  It  is  true  that  members  liquidated 
investments  to  the  amount  of  about  2  billions,  counting  from  the  peak  in 
1936  to  the  end  of  September  1937 — all  other  loans  reached  a  peak  in 
October  and  then  declined — and  that  this  naturally  weakened  the  market 
of  governments.  But  this  was,  as  our  narrative  amply  shows,  not  due  to 
strain,  nor  did  it  cause  any.  It  was,  moreover,  what  happens  in  every 
prosperity  without  producing  any  depression. 

In  deference  to  the  prevailing  belief  in  the  gratia  efficiens  of  the  supply 
of  money,  those  steps  toward  normalization  of  money-market  condi- 
tions were  retraced  precipitously.  Both  Board  and  Treasury  went  to 
entirely  unnecessary  lengths  in  the  opposite  direction  when  the  slump  had 
developed.  While  it  is  not  correct  to  say  that  what  monetary  manage- 
ment has  proved  itself  able  to  learn  was  how  to  create  slumps,  it  is  correct 
to  accept  that  panicky  retreat  as  an  indication  of  the  practical  value  of 
brakes,  the  application  of  which  so  quickly  results  in  excess  of  speed.  The 
Treasury  first  desterilized  300  millions  of  gold  in  September,1  then  reduced 
the  sterilization  policy  to  a  shadow  by  the  decree  of  Feb.  14,  1938,  and 
finally  reversed  it  by  releasing  on  Apr.  14,  1938,  the  whole  of  the  1.4 
billions  of  gold  in  the  inactive  account,  i.e.,  practically  by  transferring  the 

1  This  might  have  been  enough  to  put  a  stop  to  divestment  by  member  banks,  since 
it  taught  them  that  benevolent  powers  would  always  step  in  to  relieve  them  of  any  danger, 
however  remote,  of  illiquidity.  As  a  matter  of  fact,  however,  the  discontinuance  of  sales 
of  governments  by  members  which  then  occurred,  also  coincides  with  a  beginning  decline 
in  loans,  which  then  fell  by  about  %  billion  within  the  year.  Reversal  of  the  downward 
tendency  in  investment  was  as  much  contingent  upon  this  as  upon  the  release  of  gold,  or 
more  so. 


THE  WORLD  CRISIS  AND  AFTER  1031 

equivalent  to  the  Treasury  account  with  federal  reserve  banks.  There- 
upon, the  Board,  which  had  already  authorized  further  open-market 
operations— kept  within  modest  limits,  however — and  reduced  margin 
requirements  on  security  transactions  in  November,  also  reduced  reserve 
requirements  by  13.25  per  cent1  on  Apr.  15,  1938.  Members*  total 
reserves  having  risen  toward  the  end  of  1937  because  of  that  release  of 
those  300  millions  of  gold,  the  increase  in  monetary  silver,  the  open- 
market  operations,  and  later  also  because  of  some  purchases  of  gold, 
excess  reserves  were  more  rapidly  built  up  again.  They  were  2.5  billions 
by  Apr.  20,  1938;  2.75  for  June;  and  went  to  3.15  in  July. 

Treasury  redemption  of  bills — which  was  the  chief  use  so  far  made  of 
the  formerly  inactive  gold — sent  their  yields  to  practically  zero.  Adjusted 
demand  deposits  moved  in  the  opposite  direction  to  total  loans  plus  invest- 
ment, increasing  by  768  millions,2  while  the  latter  fell  by  250  in  the  second 
quarter  of  1938  and  cash  piled  up.  These  measures  were  more  recently 
supplemented  by  advance  on  the  other  line  of  expansionist  policy,  viz., 
by  new  rules  for  the  lending  practice  of  member  banks,  calculated  to 
"liberalize"  it.  Thus  we  have  before  us,  untouched  by  previous  experi- 
ence but  implemented  by  much  more  powerful  tools,  the  main  ideas  of  the 
policies  of  1933-1934.  No  doubt,  argument  based  on  the  logic  of  post 
hoc  ergo  propter  hoc  will  be  heard  and  read  very  soon.  It  should  be 
noticed,  moreover,  that  some  of  the  steps  taken  cannot  be  retraced,  that 
the  rest  can  be  retraced  only  with  difficulty,  and  that,  as  we  have  put  it, 
such  a  policy,  while  ineffective  in  depression,  tends  to  become  viciously 
effective  afterward.  Monetary  policy  per  se  may,  hence,  become  a  major 
factor  in  the  near  future;  but  it  had  but  little  to  do  with  the  prosperity 
of  1935  to  1937,  and  nothing  with  the  subsequent  slump. 

Finally,  we  will  return  to  income  generation  by  means  of  government 
expenditure  and  try  to  appraise  the  effects  of  its  cessation  in  1937.  Our 
historical  survey  (Chaps.  VI  and  VII)  has  supplied  us  with  instances  of 
crises  that  occurred  near  the  upper  turning  point  of  a  Juglar.  Although 
recession  is  not  depression,  the  transition  from  prosperity  to  recession, 
implying  as  it  does  a  difficult  reorientation,  always  creates  some  danger  of 
breakdown.  We  have,  it  is  true,  also  seen  that  severe  slumps  at  or  near 
that  point  are  without  exception  associated,  not  simply  with  excesses  of 
speculation,  but  with  excesses  of  speculation  induced  by  a  supernormally 
rapid  pace  of  industrial  development,  conditions  which  were  conspicu- 
ously absent  in  the  case  before  us.  But  the  idea  suggests  itself  neverthe- 

1  Reserves  against  time  deposits  were  reduced  from  6  to  5  per  cent,  against  demand 
deposits  to  22.75  per  cent  for  central  city  banks,  to  17.5  per  cent  for  reserve  city  banks, 
and  to  12  per  cent  for  country  banks. 

2  Velocity,  of  course,  fell  rapidly  and  by  the  middle  of  1938  was  about  25  per  cent  below 
its  maximum  annual  figure  (1936). 


BUSINESS  CYCLES 

ess,  that  government  expenditure  may  have  played  the  role  which  in  the 
ast  belonged  to  the  expenditure  by  innovating  firms  and  that,  consider- 
ng  its  quantitative  importance,  its  cessation  acted  in  much  the  same 
manner  as  the  cessation  of  the  latter  did  in  previous  cases.  The  elements 
of  truth  in  this  argument  are  no  more  obvious  than  its  limitations. 

There  can  be  no  doubt  not  only  that  income  generation  by  government 
must  always  create  problems  of  adaptation,  but  also  that  in  this  case  its 
timing  was  singularly  infelicitous.1  Its  high-water  mark  oame  exactly  at 
the  time  when  the  economic  process  could  most  easily  have  done  without  it 
and  its  cessation  exactly  at  the  time  when  the  economic  process  was  in  its 
most  sensitive  phase.  Reference  to  our  experimental  count  will  illustrate 
this  drastically.  The  widespread  opinion  that  the  cessation  of  that 
income  generation  was  the  "cause"  of  the  slump  thus  derives  some  sup- 
port from  our  analysis,2  although  it  should  be  superfluous  to  add  that  no 
argument  for  permanent  deficits  follows  from  that. 

But  the  explanatory  value  of  the  shock  thus  imparted  to  the  system 
should  not  be  exaggerated.  Since  no  excessive  expansion  or  speculation 
had  been  called  forth  by  the  preceding  spending  policy — any  more  than 
by  such  "natural"  prosperity  as  there  was — and  since  there  was,  as  we 
have  seen,  no  monetary  strain,  the  usual  starters  of  downward  spirals 
were  not  operative.  The  injections  were  not  suddenly  discontinued  but 
tapered  off  gradually.  Hence,  the  analogy  with  previous  crises  that 
occurred  around  the  upper  turning  point  fails  after  all.  The  obvious 
inference  is  that  the  collapse  induced  by  the  cessation  of  income  genera- 
tion was  so  severe,  and  that  the  jolt  which  it  would  in  any  case  have  been 
natural  to  expect  turned  into  a  slump,  because  under  the  surface  watered 
by  income  generation  the  processes  characteristic  of  recession  in  our  sense 
(see  Chap.  IV)  failed  to  work  as  they  always  had  worked  before.  The 
parachute  refused  to  unfold. 

5.  a.  This  view,  which  implies  the  presence  of  an  additional  and  more 
fundamental  problem,  is  hardly  controversial  and,  in  fact,  shared  by 
very  many  fellow  economists  who,  as  the  reader  knows,  offer  in  explana- 

1  So  was  the  disposal  of  the  amount  spent.     Sums  were  thrown  about  in  the  country 
almost  at  random,  without  systematic  regard  to  existing  structures  and  probable  develop- 
ments, thus  creating  industrial  and  commercial  positions  that  rested  on  nothing  but  this 
temporary  flood  of  money  and  were  bound  to  die  off  when  it  ebbed. 

2  Since  the  payments  under  the  Social  Security  Act  materially  helped  in  (slightly  more 
than)  balancing  the  combined  federal  cash  account,  it  is  even  true  to  say,  as  has  been 
said  by  some  economists,  that  those  payments  had  for  the  time  being  a  "deflationary 
effect."     As  a  long-run  proposition  this  would,  of  course,  not  be  true.     Another  point 
should  be  noticed  in  passing.     The  downturn  of  1937,  following  upon  that  of  1930,  impinged 
upon  a  business  community  which  for  the  time  being  was  supernormally  "crisis  conscious." 
Moreover,  many  concerns  may  have  harbored  vivid  recollections  of  what  the  "stand" 
they  had  made  in  1930  had  cost  them. 


THE  WORLD  CRISIS  AND  AFTER  1083 

tion  the  theory  of  vanishing  investment  opportunity.1  The  vogue  that 
this  theory  enjoys  in  this  country  is  obviously  due  to  the  occurrence  of 
that  slump  after  what  is  universally  recognized  as  an  abnormally  weak 
prosperity  (or  "recovery").  Since  in  order  to  understand  the  economic 
situation  of  our  time — and  much  besides — it  is  essential  to  realize  fully 
why  that  explanation  cannot  be  accepted,  we  will  restate  the  case  against 
it  at  the  risk  of  repetition. 

The  validity  of  that  theory  is  not  denied  on  the  ground  that  its  basic 
proposition  is  wrong.  This  proposition  can  in  terms  of  our  analysis  be 
rendered  as  follows.  Capitalism  is  essentially  a  process  of  (endogenous) 
economic  change.  Without  that  change  or,  more  precisely,  that  kind 
of  change  which  we  have  called  evolution,  capitalist  society  cannot  exist, 
because  the  economic  functions  and,  with  the  functions,  the  economic 
bases  of  its  leading  strata — of  the  strata  which  work  the  capitalist  engine 
— would  crumble  if  it  ceased:  without  innovations,  no  entrepreneurs; 
without  entrepreneurial  achievement,  no  capitalist  returns2  and  no 
capitalist  propulsion.  The  atmosphere  of  industrial  revolutions — of 
"progress" — is  the  only  one  in  which  capitalism  can  survive.  Hence 
the  capitalist  organism  cannot,  in  case  opportunities  for  innovations  give 
out,  settle  down  into  a  stationary  stage  without  being  vitally  affected, 
as  it  could  if  "changes  in  production  functions"  were  an  incident  to  its 
life  process  and  not  the  essence  of  it.  In  this  sense  stabilized  capitalism 
is  a  contradiction  in  terms.  Moreover,  it  is  reasonable  to  expect  that 
this  kind  of  stabilization  would  produce  a  class  of  abnormalities  and 
instabilities  of  its  own.  There  would  be  increasing  reluctance  to  invest 
or  even  reinvest,  a  tendency  to  "live  on  capital,"  to  hold  on  to  balances, 
to  recreate  vanishing  returns  by  all  the  shifts  open  to  a  class  which, 
though  by  then  economically  functionless,  yet  would,  like  its  feudal 
predecessor,  for  a  time  retain  the  powers  acquired  by  and  associated  with 
the  functions  previously  filled.  Maladjustments,  unemployment  and 
underutilization  of  resources — though  now  of  a  different  nature — and 
neutral,  unstable,  and  subnormal  equilibria  might  hence  well  stay  with  a 
nonexpanding  world.3 

1  For  an  admirable  exposition  of  that  theory,  see  A.  H.  Hansen,  Full  Recovery  or 
Stagnation,  1938. 

2  This,  of  course,  is  putting  it  strongly.     But  the  necessary  qualifications  must  be 
supposed  to  be  by  now  familiar  to  the  reader,  as  must  also  the  argument  summed  up  in  that 
statement.     Jt  should  be  observed  that  what  we  have  called  Growth  is  not  separately 
mentioned.     The  reasons  for  believing  that  the  class  of  phenomena  thus  designated  is  not 
important  for  the  argument  which  is  to  follow  will  presently  become  obvious. 

8  Such  a  world  would  not  display  cycles  in  our  sense:  a  cycle  in  a  nonexpanding  world 
is  also  a  contradiction  in  terms.  But  fluctuations  of  the  type  which  we  have  called  waves 
of  adaptation  (Chap.  IV)  would  continue  for  some  time.  And  the  circumstances  alluded 
to  in  the  text  may,  while  transition  lasts,  also  cause  fluctuations  of  yet  another  type. 


1034  BUSINESS  CYCLES 

The  colors  of  this  picture  will  fade  if  proper  account  be  taken  of  the 
fact  that  transition  to  a  stationary  state  would  not  be  sudden  but  would 
necessarily  come  about  by  slow  degrees  (see  below).  Also,  it  must  be 
borne  in  mind  that  the  proposition  in  question  is  not  as  a  rule  formulated 
in  that  way.  Some  writers  try  to  demonstrate  it  by  means  of  models 
which  assume  unchanging  production  functions  or  "methods  of  produc- 
tion "  and  thus  exclude  the  pivot  on  which  it  turns.  Most  writers  unduly 
stress  the  mere  mechanics  of  the  saving-investment  process.  But  as  far 
as  the  result  is  concerned,  there  is  nevertheless  affinity  between  their 
view  and  the  one  submitted  in  this  book.  We  may  even  admit  that  one 
of  the  difficulties  of  that  transition  may  proceed  from  people  wanting  to 
invest  and  getting  ready  to  invest,  while  they  are  not  able  to  do  so  at  rates 
of  returns  acceptable  to  them  (see  below).  It  will,  therefore,  be  con- 
venient to  accept  the  current  slogan  for  the  purposes  of  the  argument  in 
hand.  The  connecting  link  which  allows  us  to  do  so  is  the  fact  that 
innovation  is,  directly  and  indirectly,  the  great  source  of  investment 
opportunity. 

Nor  do  we  take  issue  with  the  companion  proposition  that  investment 
opportunity  in  this  sense  may,  and  in  fact  is  quite  likely  to,  vanish  some- 
time in  the  future.  The  reasons  usually  offered  for  this  are  old  acquaint- 
ances of  ours.1  For  instance,  although  we  hold  that  the  conquest  of  the 
air  may,  entrepreneurially  speaking,  be  as  important  as  or  more  important 
than  the  conquest  of  India  and  that,  from  the  standpoint  of  our  analysis, 
the  two  are  exactly  the  same  kind  of  thing,  we  shall  not  deny  that  oppor- 
tunities of  the  latter  type  are  being,  or  ultimately  will  be,  exhausted. 
Or,  although  we  hold  that  nothing  at  all  reliable  can  be  predicated  about 
it,  we  do  not  deny  the  possibility  that  technological  innovations  may 
some  day  give  out,  either  "objectively"  or  because  people  do  not  care  to 
proceed  with  the  available  ones.2  We  have  even  an  element  of  our  own 
to  add.  The  mechanization  of  "progress"  (Chap.  Ill)  may  for  entre- 
preneurs, capitalists,  and  capitalist  returns  produce  effects  similar  to 
those  which  cessation  of  technological  progress  would  have.  Even  now, 
the  private  entrepreneur  is  not  nearly  so  important  a  figure  as  he  has  been 
in  the  past.  We  have,  moreover,  noticed  (Chaps.  VII  and  XIV)  the 
implications  of  chemical  and  other  developments  which  may  result  in 
making  innovation  capital  saving  or  at  least  less  capital  absorbing  than, 
say,  it  has  been  in  the  railroad  age.  Also,  it  may  well  be  true  that  an 

1  See,  in  particular,  Chaps.  Ill,  IV,  and  IX  (retardation). 

2  It  should  be  repeated,  however,  (see  Chap.  IX)  that  any  assertion  about  this  would 
be  pure  speculation.     So  would  any  assertion  to  the  effect  that  the  innovations  which 
have  materialized  so  far  were  more  important,  more  profitable,  or  more  capital  absorbing 
than  those  that  are  to  come.     Such  assertions  repeat  Malthus's  methodological  error  with  a 
vengeance. 


THE  WORLD  CRISIS  AND  AFTER  1035 

increasing  proportion  of  the  "things  still  to  be  done"  will  lend  itself  to 
public  rather  than  private  enterprise,1  although  this  would  per  se  not 
mean  more  than  the  addition  of  another  component  to  a  tendency 
toward  public  enterprise  which  exists  independently  of  it. 

Finally,  we  do  not  even  exclude  the  possibility  that  investment  oppor- 
tunity might  vanish  through  saturation.  The  argument  from  declining 
birth  rates,  in  particular,  loses  but  little  by  the  fact  that  it  is  often  inade- 
quately formulated.  Reduced  "need"  for  expanding  capital  equipment 
in  a  society  in  which  population  increases  at  a  decreasing  rate — still  more 
in  a  society  in  which  it  is  stationary  or  declining — is  not  the  point.  It 
does  not  matter  whether  the  purposes  to  be  served  by  an  expansion  or 
reorientation  of  the  productive  apparatus  strike  the  observer  as  partic- 
ularly "necessary"  or  not — whether  it  is  radio  sets  or  cradles  that  are  in 
demand.  And  so  long  as  the  majority  of  people  in  the  civilized,  let  alone 
the  uncivilized,  nations  are  as  far  from  anything  like  saturation2  as  they 
are  now,  no  shrinkage  of  total  investment  opportunity  will  result  from 
the  saturation  of  a  particular  want — even  if  we  disregard  the  temporary, 
but  for  the  time  being  very  important,  effect  of  the  changing  age  distribu- 
tion. But  there  is  a  more  fundamental  objection  which  at  this  state  of 
our  analysis  should  be  familiar  to  the  reader:  "needs,"  whatever  they 
may  be,  are  never  more  than  conditioning  factors,  and  in  many  cases 
mere  products  of  entrepreneurial  action;  it  is  not  they  that  set  the  capital- 
ist engine  into  motion,  as  the  old  household  examples  (China  and  so  on) 
show;  and  economic  development  (capital  consumption  included)  has 
never  been  conspicuous  in  the  countries  which  to  the  observer  seem  to  be 
most  lavishly  supplied  with  needs.  However,  the  argument  may  be 

1  The  criteria  according  to  which  it  is  to  be  decided  whether  a  given  proposition  is 
economically  more  fitted  for  public  than  for  private  enterprise  vary,  of  course,  historically 
and  are  at  present  blurred  by  the  prevailing  preference  for  the  former,  which  is  largely 
extra-economic.     But  we  need  not  go  into  this  beyond  reminding  ourselves  that  there 
is  a  class  of  cases  in  which  inadequacy  of  private  profits  indicates  inadequacy  of  social 
advantage. 

2  Once  more  we  meet  that  overemphasis  on  the  possibilities  of  expansion  within  existing 
lines  which  we  repeatedly  met  before.     But  it  should  be  observed  that  even  these  are 
important  enough  to  negative  any  idea  of  deadlock  from  that  source  for  quite  a  time. 
Another  argument  may  be  noticed  here.     It  has  been  held  that  the  process  of  providing 
"capitalist"  equipment  for  the  population — the  stock  of  producers'  goods  necessary  for 
the  Boehm-Bawerkian  "roundaboutness" — was  a  historically  unique  and  uniquely  capital- 
absorbing  task,  which  has  been  performed  once  for  all  in  the  nineteenth  century.     This 
seems  not  only  to  assume  that,  within  the  production  functions  that  existed  at  any  time, 
capital  equipment  has  been  carried  to  saturation  point,  but  also  to  overlook  that  the  inser- 
tion of  new  production  functions  in  many  cases,  not  to  say  typically,  annihilates  the  old 
equipment  economically,  so  that  that  task  has  to  be  periodically  repeated  and,  as  experience 
abundantly  shows,  repeated  by  means  of  new  savings  and  new  credit  creation.     Long- 
distance transportation  by  motorcar,  "replacing"  long-distance  transportation  by  railroads, 
is  surely  not  financed  from  the  depreciation  accounts  of  the  latter. 


1036  BUSINESS  CYCLES 

upheld,  at  least  to  some  extent,  in  another  way:  as  has  been  pointed  out, 
provision  for  an  indefinite  family  future  is  of  central  importance  in  the 
scheme  of  bourgeois  motivation,  and  much  driving  power  may  be  elimi- 
nated by  childlessness. 

b.  But  we  do  take  issue  with  the  third  proposition,  which  asserts  the 
relevance  of  those  considerations  for  the  diagnosis  of  the  situation  of 
1938.  Obviously,  we  have  been  speaking  of  longest-run  tendencies  just 
now.  Opportunities  for  technological  or  organizational  qr  commercial 
innovation  cannot  be  thought  of  as  vanishing  (if  they  are  vanishing) 
except  very  gradually.  If  there  actually  be  a  general  tendency  toward 
decline  in  capital  absorption,  it  can  assert  itself  only  in  time,  though 
shocks  to  individual  industries  may  be  both  sudden  and  serious.  The 
rate  of  increase  in  population  declines  imperceptibly  per  year.  The 
call  of  entrepreneurial  adventure  is  too  deep-seated  to  cease  dramatically 
of  itself.  And  so  on.  Such  tendencies,  even  if  well  established  as  some 
of  them  undoubtedly  are,  qualify  but  ill  for  the  task  of  explaining  the 
peculiarities  of  a  particular  Juglar.  They  may  affect  contour  lines  over 
time  and  bend  them  downward.  But  they  cannot  explain  the  weakness, 
relatively  to  its  predecessor,  of  any  given  prosperity,  and  they  look 
absurd  in  the  role  of  explaining  factors  of  a  sudden  slump.  If  it  be  held, 
nevertheless,  that  one  or  all  of  them  suddenly  acquired  dominating 
importance  at  any  given  historical  juncture,  such  an  assertion  requires, 
in  order  to  be  taken  seriously,  proof  not  only  of  a  secular  tendency  or 
"trend"  but  of  the  presence  of  circumstances  adequate  to  account  for  so 
improbable  an  occurrence  as  sudden  action  would  be. 

It  has  been  observed  above  that  the  essentially  gradual  modus  operandi 
of  those  tendencies  must  enter  into  any  speculations  about  the  phenomena 
to  be  expected  from  vanishing  investment  opportunities.  For  instance, 
there  is  no  warrant  for  the  assumption  which  has  been  made  the  basis  of 
far-reaching  conclusions,  that  in  the  face  of  them  people  will  go  on  saving 
at  a  rate  sufficient  to  produce  difficulties.  Owing  to  the  persistence  of 
habits,  this  could  conceivably  be  so  in  the  short  run  of  a  depression  phase, 
though  we  have  seen  plenty  of  reasons  to  doubt  it.  But  in  the  short  run 
investment  opportunity  cannot  decline  perceptibly.  And  in  the  long 
run  there  is  no  reason  to  suppose  that  savings — and  both  things  and 
psyches  in  general — will  not  adapt  themselves.  Moreover,  it  is  obvious 
not  only  that  declining  rates  of  savings  will  accompany  declining  birth 
rates  because  both  phenomena  flow  from  the  same  sociopsychological 
source,  but  also  that  there  is  a  causal  connection  between  the  two. 

But  reasons  less  general  than  that  estop  us  from  accepting  the  theory 
in  question.  Whatever  may  be  thought  about  those  "trends"  and  the 
way  in  which  they  operate,  none  of  them  has  in  this  country  advanced 
far  enough  to  bear  it  out.  We  have  seen  in  some  detail  that  "objective " 


THE  WORLD  CRISIS  AND  AFTER  1037 

opportunities  are  not  lacking.  We  are  less  than  ten  years  removed  from 
as  vigorous  a  prosperity  as  was  ever  witnessed  and  from  a  depression 
provably  due,  in  the  main,  to  the  pace  of  preceding  "progress."  It  has 
been  argued  that  that  prosperity  differed  in  character  from  previous 
ones  by  the  prominence  of  (durable)  consumers'  goods  production,  and 
hence  indicated  that  a  fundamental  change  had,  already  then,  occurred 
in  the  cyclical  process.  Of  course  that  prosperity  differed  from  the  Jug- 
lar  prosperities  of  the  Kondratieff  upgrade.  But  it  did  not  differ  in 
character  from  the  comparable  Juglar  prosperities  of  the  preceding 
Kondratieff  downgrades,  and  therefore  does  not  indicate  any  fundamental 
change  in  the  working  of  the  capitalist  organism.  Expansion  of  pro- 
duction of  consumers'  goods,  including  expansion  in  the  fields  of  utilities 
and  of  public  works,  was,  proportions  guarded,  no  less  prominent  a 
feature  in  the  developments  of  the  twenties,  thirties,  seventies,  and 
eighties  of  the  nineteenth  century.  Nor  can  it  be  urged  that  funda- 
mentally new  opportunities  of  first-rate  magnitude  are  not  in  prospect. 
Barring  the  question  whether  that  is  so,  it  is  sufficient  to  reply  that  in 
the  eighteen-twenties  hardly  anybody  can  have  foreseen  the  impending 
railroad  revolution  or,  in  the  eighteen-seventies,  electrical  developments 
and  the  motorcar.  No  less  an  authority  than  John  St.  Mill1  compromised 
himself  by  holding  in  1870  that  the  possibilities  of  capitalist  enterprise 
were  substantially  exhausted.  Vestigia  terreant. 

As  applied  to  the  American  situation  of  today  and  to  the  abnormalities 
of  the  current  Juglar,  the  theory  that  the  capitalist  process  is  stagnating 
from  internal  causes  inherent  to  its  logic  and  that  income  generation  by 
government  is  nothing  but  the  self-defense  of  a  shriveling  organism,  is 
therefore  a  complete  misfit — at  best  a  mistaken  interpretation  of  certain 
aftereffects  of  the  world  depression,  at  worst  the  product  of  wishful  think- 
ing on  the  part  of  all  those  who  crave  for  a  presentable  basis  for  policies 
they  approve.  It  still  retains  two  merits,  however.  The  one  con- 
sists in  the  many  elements  of  truth  which,  as  we  have  seen,  enter  into  its 
arguments  as  distinguished  from  its  application.  The  other  consists  in 
the  recognition,  by  implication  at  least,  of  the  fact  that,  as  any  social 
system  depends  for  its  functioning  and  survival  on  the  actual  delivery  of 
the  premia  it  holds  out,  so  capitalism  depends  for  its  functioning  and  sur- 
vival on  the  actual  delivery  of  the  returns,  anticipation  of  which  provides 
its  motive  power.2  For  this  is,  after  all,  what  the  stressing  of  investment 

1  This  has  been  pointed  out  to  the  writer  by  Mr.  R.  Abel-Musgrave. 

*  See  Chap.  XIV,  Sec.  C,  2.  This  would  in  the  times  of  intact  capitalism  have  been 
taken  for  granted.  In  our  time,  however,  the  attempt  to  run  capitalism  in  an  anticapitalis- 
tic  way  has  given  rise  to  arguments  which  came  near  to  denying  it.  Hence  the  recognition 
by  the  theory  of  vanishing  investment  opportunity  of  that  rather  obvious  point  may  well 
be  recorded  to  its  credit. 


1038  BUSINESS  CYCLES 

opportunity  amounts  to.  Slightly  extending  and  modifying  the  mean- 
ing of  that  phrase,  we  may  hence  continue  to  use  it  ourselves  and 
agree  that  it  is  vanishing  investment  opportunity  which  is  the  matter 
with  present-day  capitalism — anything  can,  in  fact,  be  put  into  that 
form,  the  structural  principles  of  the  capitalist  process  being  what 
they  are.  And  our  task  then  reduces  to  substituting  for  unconvincing 
reasons  why  investment  opportunity  should  be  vanishing,  a  more 
convincing  one. 

c.  The  analysis  of  Chap.  XIV,  Sec.  B  supplies  it:  capitalism  produces 
by  its  mere  working  a  social  atmosphere — a  moral  code,  if  the  reader  pre- 
fer— that  is  hostile  to  it,  and  this  atmosphere,  in  turn,  produces  policies 
which  do  not  allow  it  to  function.  There  is  no  equilibrating  apparatus  to 
guarantee  that  this  atmosphere  or  these  policies  should  develop  in  such 
a  way  as  to  prevail  in  the  fullness  of  time,  i.e.,  when  the  capitalist  process 
will  have  really  spent  its  force  or  be  spending  it.  Whenever  they  prevail 
sooner,  there  is  danger  of  a  deadlock,  by  which  we  mean  a  situation  in 
which  neither  capitalism  nor  its  possible  alternatives  are  workable.  This 
is  what,  to  a  certain  extent  and  presumably  not  yet  for  good,  has  happened 
in  this  country. 

It  might  be  replied  that  anticapitalist  attitudes  are  also,  like  the 
tendencies  adduced  by  the  theory  of  vanishing  investment  opportunity 
in  its  usual  acceptance,  a  matter  of  slow  growth  and,  hence,  similarly 
open  to  one  of  the  objections  raised  against  that  theory  above.  But  we 
are  able  to  do  in  this  case  what  cannot  be  done  for  those  tendencies,  viz., 
to  show  that,  and  how,  that  attitude  came  suddenly  to  a  head  and 
suddenly  acquired  dominating  importance.  And  anticapitalist  policies, 
unlike  attitudes,  may  be  dated.1  The  coincidence  in  time  between 
them  and  disappointing  performance  of  the  economic  engine  is  indeed 
striking.  We  will  survey  them  under  the  headings  of  Fiscal,  Labor,  and 
— for  want  of  a  better  expression — Industrial  policies. 

At  least  since  1932  the  burden  of  direct  taxation  imposed  upon  that 
part  of  the  national  revenue  which  goes  to  the  higher  and  highest  brackets 

1  It  should  be  observed  at  once  that,  anticapitalist  measures  being,  of  necessity,  meas- 
ures hostile  to  private  investment  opportunity,  the  accredited  exponents  of  the  theory  of 
vanishing  investment  opportunity  must  perforce  agree  with  the  argument  that  is  to  follow; 
for  the  consequences  of  inadequate  investment  opportunity  are  obviously  independent  of 
its  causes.  They  will  be  the  same  whether  these  causes  are  internal  or  external,  i.e., 
whether  the  process  itself,  by  virtue  of  the  law  of  its  life,  produces  inadequate  margins, 
or  these  margins,  if  produced,  are  or  would  be  taken  from  recipients,  or,  finally,  the  anticipa- 
tion of  them  is  in  other  ways  prevented  from  having  its  normal  effect.  Hence,  those 
economists  will  in  any  case,  even  if  unconvinced  by  our  argument  against  their  explanation 
of  vanishing  investment  opportunity,  at  least  have  to  insert  ours  into  their  schema.  If 
they  do  insert  it  in  a  place  appropriate  to  the  importance  of  its  constituent  elements,  there 
will  not  be  much  room  left  for  difference  of  opinion. 


THE  WORLD  CRISIS  AND  AFTER  1039 

was  undoubtedly  high  enough1  to  affect  "subjective"  investment 
opportunity  or,  as  we  have  previously  expressed  it,  to  shift  the  watershed 
between  "to  do  and  not  to  do."  No  other  than  direct  or  mechanical 
effects  need,  however,  be  attributed  to  this  burden  until  roughly  1933- 
1934,  because  the  increase  in  taxation  was  then  accepted  as  a  sacrifice  to 
be  made  in  a  national  emergency,  as  it  had  been  during  the  war.  But 
from  the  revenue  act  of  1934  on,  this  was  no  longer  so.  Permanence  of 
the  burden  for  reasons  unconnected  with  emergency,  involving  a  transfer 
or  redistribution  of  wealth  which  in  the  highest  brackets  amounted  to  the 
socialization  of  the  bulk  of  private  income,  and  in  some  cases  taxation  for 
taxation's  sake  and  regardless  of  insignificance  of  results  for  the  Treasury,2 
then  became  part  of  an  established  policy,  the  general  drift  of  which  was 
not  reversed  in  1938.  Some  outlines  of  the  theory  of  the  subject  have 
been  presented  in  the  preceding  chapter.  Aspects  other  than  effects  on 
the  process  of  economic  evolution  are  irrelevant  to  our  purpose.  The 
quantitative  importance  of  the  change  to  the  interests  concerned  is 
unquestionable  and  unquestioned.  Hence,  we  need  not  go  into  details  or 
follow  up  the  successive  steps  embodied  in  the  revenue  acts  from  1934  to 
1937,  but  can  confine  ourselves  to  the  following  comments. 

As  the  above  suggests,  the  writer  is  inclined  to  stress  the  importance  of 
the  income,  corporation,  and  estate  taxes  at  the  expense  of  others  which, 
being  novelties,  have  been  more  widely  discussed.  The  facts  that  the 
limit  of  exemption  from  the  income  tax  is  very  high,  the  flat  rate  very  low, 
and  the  surtax  distinctly  moderate  up  to  an  income  of  $30,000,  are  irre- 
levant to  the  argument.  It  is  above  that  range,  principally  within  a 
group  of  not  more  than  30,000  or  40,000  taxpayers  that,  the  structure  of 
American  industry  being  what  it  is,  those  taxes,  raised  within  a  few  years 
to  their  present  figures,  exert  a  serious  influence  on  "capital  supply"3  and 

1  For  a  careful  evaluation  of  that  burden,  including  an  estate  tax  equivalent  and 
covering  the  years  1924  and  1927  to  1933,  see  M.  Yaple  Sweezy,  Direct  Taxes  by  Income 
Classes,  American  Economic  Review  for  December  1936. 

2  The  insignificance  of  financial  results  is  very  striking  in  the  case  of  the  estimates — 
which  are  what  is  relevant  as  regards  purpose — of  additional  revenue  from  the  revenue 
act  of  1935.     The  increases  of  the  surtax  were,  for  example,  to  yield  45  millions  more,  the 
graduated  income  tax  37  millions,  the  excess  profits  tax  10  millions,  the  increase  of  the 
estate  tax  plus  the  gift  tax  101  millions — the  nibbling  of  a  mouse  at  the  mountain  of  the 
deficit.     The  arguments  that  this  was  a  matter  of  that  budget  and  that  in  future  booms 
much  higher  yields  could  be  expected  are  not  to  the  point.     The  latter,  moreover,  begs 
the  fundamental  question. 

3  The  writer  does  not  wish  to  stress,  under  the  circumstances  of  the  past  years,  short- 
run  effects  on  quantity  of  monetary  capital  and  its  rate  of  increase.     As  far  as  these  go, 
that  fiscal  policy  may  even  have  had  a  net  result  favorable  to  prosperity  and  unfavorable  to 
depression  through  enforcing  an  increase  in  total  expenditure.     That  element  is  primarily 
stressed  in  their  interesting  study  on  Economic  Consequences  of  Recent  American  Tax 
Policy,  Supplement  I  to  Social  Research,  by  Professors  Colm  and  Lehmann,  who  attempt 


1040  BUSINESS  CYCLES 

business  behavior,  which,  of  course,  is  greatly  intensified  by  the  failure  of 
legislation  to  permit  the  carrying  forward  of  business  losses,1  by  the  new 
treatment  of  personal  holding  companies  and  by  other  inroads  into  actual 
or  potential  capital. 

The  so-called  capital  gains  tax  has  been  held  responsible  for  having 
accentuated,  if  not  caused,  the  slump.  The  writer  is,  however,  unable  to 
see  that  it  can  have  had  much  to  do  with  the  processes  of  the  current 
Juglar  except  by  way  of  damping  speculative  ardors  and  thereby  making 
issues  of  stock  more  difficult  than  they  would  have  been.  %The  financing 
of  the  positive  phase  cannot,  considering  the  abundance  of  cheap  money, 
have  been  seriously  interfered  with  by  this;  the  subsequent  slump  should, 
if  anything,  have  been  mitigated  by  it.  Other  points,  in  particular  the 
effect  it  exerts — not  by  reducing  " over-saving"  but — by  enforcing  dis- 
saving, though  relevant  to  a  prognosis  of  the  results  to  be  expected  in  the 
future  from  the  capitalist  engine  and  not  substantially  affected  by  the 
modifications  introduced  by  the  revenue  act  of  1938,  need  not  concern  us 
here. 

The  antisaving  theories  and  the  ressentiments  of  the  day  found  a  very 
characteristic  expression  in  the  special  surtax  on  undistributed  corporate 
income  (undivided  profits  tax),  which  ranged  from  7  to  27  per  cent. 
Discarding  again  the  question  of  the  long-run  effects  which  the  measure 
might  have  had  if  it  had  been  allowed  to  remain  on  the  statute  book,  we 
may  split  immediate  effects  into  those  of  a  further  increase  in  the  burden 
on  corporate  income  and  those  of  the  specific  penalty  imposed  on  corporate 
accumulation,  and  confine  ourselves  to  the  latter.  It  possibly  resulted  in 
an  absolute  and  relative  increase  in  distributed  income  which  is  neither 
certain  nor  easy  to  evaluate  because  there  were  also  other  reasons  for  the 
increase  that  actually  occurred,  but  which,  whether  great  or  small, 
presumably  increased,  or  helped  to  counteract  the  decrease  of,  system 
expenditure.  Nevertheless,  the  measure  may  well  have  had  a  paralyzing 
influence  on  enterprise  and  investment  in  general.  The  actual  presence  of 
accumulated  "reserves,"  and  the  possibility  of  accumulating  them 
quickly,  strengthens  the  position  of  a  concern  with  respect  to  the  risks  and 
chances  of  innovation  and  expansion  which  it  confronts.  One  of  the  causes 
of  the  efficiency  of  private  business  is  that,  unlike  the  politician  or  public 

quantitative  evaluation  of  the  difference  made  to  the  supply  of  capital.  But  to  the  writer 
supply  in  the  sense  of  "willingness  to  sell,"  that  is  to  say  in  this  case,  willingness  to  invest, 
seems  to  have  been  the  more  immediately  important  thing  affected. 

1  That  feature,  for  which  no  rational  argument  has  ever  been  offered  to  the  writer's 
knowledge,  is  more  important  than  it  seems.  A  loss  which  can  be  carried  forward  without 
penalty  is  one  thing,  a  loss  which  cannot,  quite  another  thing.  A  risk  which  it  may  be 
rational  to  take  in  the  first  case  will  frequently  have  to  be  refused  in  the  latter.  This  would 
not,  of  course,  apply  to  "small"  or  even  moderate  taxes. 


THE  WORLD  CRISIS  AND  AFTER  1041 

officer,  it  has  to  pay  for  its  mistakes.  But  the  consequences  of  having  to 
do  so  are  very  different  according  to  whether  it  risks  owned  or  borrowed 
"  funds,"  or  whether  a  loss  will  only  reduce  surpluses  or  directly  impinge 
upon  original  capital.  Adequate  book  reserves  are  as  necessary  a  requi- 
site as  adequate  stocks  of  raw  material,  and  in  their  absence,  or  with 
reduced  facilities  of  acquiring  or  replenishing  them,  an  entirely  different 
and  much  more  cautious  business  policy  would  impose  itself.  In  pros- 
perity, investment  opportunities  would  be  seen  in  a  perspective  of 
reduced  proportions;  in  depression,  firms  would  have  to  bow  more  readily 
to  the  storm.  In  the  latter  case  in  particular,  the  important  class  of 
considerations — pure  business  considerations  among  them — that  used  to 
induce  many  firms  "to  make  a  stand  "  for  some  time,  even  at  considerable 
immediate  loss,  would  tend  to  vanish  from  the  businessman's  mind.  All 
this,  it  is  true,  vanishes  from  the  economist's  mind  as  soon  as  he  buries 
himself  in  the  mechanics  of  aggregative  theory.  But  many  industries, 
which  are  among  the  chief  economic  assets  of  the  nation  and  of  which  the 
automobile  industry  is  the  standard  instance,  would  under  a  regime  of 
undivided  profits  taxes  never  have  developed  as  they  did.  And  as  regards 
the  current  Juglar,  the  actual  course  of  events  both  during  prosperity 
and  during  the  slump  is  compatible  with  the  opinion  that  this  attack  on 
the  foundation  of  corporate  finance  weakened  the  former  and  intensified 
the  latter  to  a  non-negligible  extent.1 

It  should  be  observed  that  this  is  a  matter  of  value  of  assets  and  not  of 
liquidity,  which  under  the  conditions  prevailing  in  this  country  since 
1931  was  never  a  problem  for  a  concern  of  unimpaired  standing.  Simi- 
larly, the  argument  that  accumulations  make  it  easier  for  a  concern  to 
live  in  depression  and  to  "cushion"  the  effects  of  depression  on  the  eco- 
nomic process  by  keeping  up  dividend  and  wage  payments  cannot  be  met 
by  pointing  out  that  only  a  part  of  total  accumulations  is  held  in  cash  or 
near-cash  and  that  the  rest  cannot  be  "paid  out."  It  is  true  that  from 
the  standpoint  of  the  individual  management  liquidity  constitutes  an 
advantage.  The  ease  with  which  the  bulk  of  American  large-scale  indus- 
try steered  through  the  vicissitudes  of  1931  and  1932  was  to  a  consider- 
able extent  due  to  it.  It  is  also  true  that  accumulations  which  are  held  in 
a  liquid  form  tend  to  work  in  an  anticyclical  sense.  But  this  must  not  be 

1  An  exponent  of  antisaving  views  disposes  of  that  opinion  as  "ballyhoo.**  After 
recording  his  belief  that  the  country  does  not  want  additional  investment  and  that,  if  it 
did,  funds  could  easily  be  secured  by  borrowing,  he  asks:  "Why  did  it  (the  undivided  sur- 
plus tax)  not  produce  this  sad  effect  (the  slump)  sooner?"  No  serious  economist  has,  to 
the  writer's  knowledge,  held  that  the  slump  was  solely  caused  by  this  tax.  But  if  timing 
of  effects  carries  any  weight  (the  reader  knows  that  we  attach  but  limited  importance  to 
time  sequences),  it  is  nothing  short  of  ideal  in  this  case.  The  act  went  into  force  in  June 
1986.  Its  effects  were  due  to  show  themselves  urbi  et  orbi  in  the  first  quarter  of  1987. 
And  at  the  end  of  the  second  the  first  symptoms  of  impending  difficulties  appeared. 


1042  BUSINESS  CYCLES 

confused  with  the  point  which  the  writer  is  trying  to  make  and  which,  in 
this  case,  is  entirely  independent  of  cash  considerations,  though  it  would 
not  be  so  for  other  times  and  countries. 

The  effect  on  the  combined  federal  cash  account  of  the  method  chosen 
for  financing  the  social  security  program  has  been  mentioned  before.  No 
further  attention  need,  hence,  be  paid  to  the  money-market  and  expendi- 
ture aspects  of  the  payments  into  the  Old  Age  Reserve  Account  and  the 
Unemployment  Trust  Fund.  Independently  of  this,  the  tax  on  pay  rolls 
levied  on  firms  was  of  course  a  nonnegligible  element  in  the'increase  of  the 
total  fiscal  burden  which  occurred  in  1937.1  The  question  of  effects  raises 
difficult  problems  in  transference.  In  a  situation  in  which  wage  rates  are 
firmly  upheld  and  prices  of  the  products  of  "big  business"  not  allowed 
to  rise,  increase  of  the  tax  to  the  full  amount  ultimately  contemplated 
may  not  only  produce  additional  unemployment,  but  also  be  sufficient, 
as  comparison  with  corporate  net  earnings  shows,  to  cause  paralysis  in 
some  industries,  such  as  would,  for  example,  enforce  the  creation  of 
another  and  much  more  stringent  NRA.  But  for  the  time  being  no  major 
effect  can  be  attributed  to  this  tax  taken  by  itself. 

Labor  policies  reduced  investment  opportunity — besides  employment 
per  unit  of  output — mainly  by  forcing  up  wage  rates.  We  have  seen, 
however,  that  not  all  of  the  increase  which  actually  occurred  can  be  attrib- 
uted to  those  policies;  and  precisely  because  rising  rates  were  to  a 
considerable  extent  met  by  labor-saving  rationalization,  the  effect  on 
investment  opportunity  was  presumably  not  very  great.2  Costs  incident 
to  employing  labor  were  also  increased  in  other  ways.  And  more  than 
elsewhere  it  is  here  necessary  to  define  investment  opportunity  widely  and 
to  take  account  of  the  less  measurable  effects  of  increasing  difficulties  in 
operating  plants  which  the  growth  of  a  new  body  of  administrative  law 
entails.  A  major  measure  of  this  kind,  the  National  Labor  Relations  Act 
(July  5,  1937)  was  placed  on  the  statute  book  in  the  period  under  survey. 
As  the  reader  knows,  more  vigorous  use  was  immediately  made  of  the 

1  The  steel  industry,  for  instance  (American  Iron  and  Steel  Institute;  identical  concerns), 
paid  for  1937  156.6  millions  in  federal,  state,  and  local  taxes,  or  about  40  per  cent  of  net 
earnings  or  15  per  cent  of  pay  rolls,  or  about  60  per  cent  more  than  in  the  preceding  year. 
About  37  of  the  60  per  cent  are  accounted  for  by  the  increase  in  the  pay  roll  tax. 

2  It  will  be  recalled  that  as  far  as  meeting  rising  wages  in  that  way  involves  additional 
outlay  on  plant  and  equipment,  investment  opportunity  may  temporarily  be  increased. 
Such  facts  as  that  in  the  steel  industry  the  composite  price  of  products  was  in  July  1938  a 
little  below  the  1923  to  1929  average,  while  average  hourly  money  earnings  of  labor  were 
higher  by  33  per  cent  must  be  interpreted  in  this  light.     It,  nevertheless,  remains  true 
that  the  course  of  wages  was  one  of  the  factors  that  account  for  that  industry's  failure  to 
repair  its  damaged  financial  structure,  as  evidenced  by  the  fact  (Iron  and  Steel  Institute) 
that  at  the  end  of  1937  its  surplus  and  reserves  were  still  but  little  more  than  half  of  what 
they  had  been  in  1920. 


THE  WORLD  CRISIS  AND  AFTER  1043 

facilities  created  by  it  than  is  suggested  by  its  actual  contents,  which 
keep  within  the  most  ordinary  lines  of  labor  policy  in  modern  democracies 
and  only  develop  the  principles  of  earlier  legislation,  such  as  the  labor 
clauses  of  railroad  acts,  the  Clayton  Act,  certain  acts  passed  during  the 
war,  Sec.  7a  of  the  recovery  act,  and  other  enactments.  Official  support 
given  to  the  campaign  of  the  Committee  for  Industrial  Organization  and 
lending  to  the  act  a  color  not  naturally  its  own,  must  be  listed  inde- 
pendently. But  after  the  fullest  allowance  for  these  and  other  elements 
of  the  case,  we  shall  still  be  left  with  the  result  that  labor  policies — more 
precisely,  what  has  actually  been  done  in  the  field  of  labor  policy — were 
not,  taken  by  themselves,  of  decisive  importance  in  shaping  the  business 
situations  of  those  years. 

As  regards  what  we  have  called  industrial  policies  unfavorable  to 
investment  opportunity — or,  what  economically  amounts  to  the  same 
thing,  to  action  being  taken  on  any  "objective"  investment  opportunity, 
whether  declining  or  not,  that  may  have  been  present — two  instances  will 
sufficiently  illustrate  what  we  mean.  First,  we  have  seen  reasons  to 
expect  that  developments  in  the  field  of  public  utilities  would  be  a  leading 
feature  of  the  current,  as  they  had  been  in  the  preceding  Juglar.  We 
have  also  seen  that,  barring  federal  enterprise,  that  expectation  was  not 
fulfilled.  The  writer  does  not  see  how  it  could  possibly  be  denied  that  in 
this  case  existing  investment  opportunity  was  prevented  from  having  its 
normal  effect,  not  so  much  by  what  was  actually  done,  but  by  the  blanket 
threat  behind  it.  Expected  competition  from  federal  or  municipal  power 
plants  was  a  factor  in  some  sectors.1  The  Public  Utilities  Holding  Com- 
pany Act  endangered  the  American  solution  of  the  fundamental  problem 
of  power  finance.  But  the  decisive  element  of  the  situation  was  that 
indefinite  threat:  executives  and  investors  would  have  had  to  be  com- 
pletely blind  to  the  political  forces  that  were  being  marshaled  against 
them,  if  they  had  been  prepared  to  take  the  responsibility  for,  or  to  cooper- 
ate in,  new  investment  on  a  large  scale.  The  case  thus  serves  to  show  not 
only  how  unrealistic  any  theory  of  investment  opportunity  is  which  leaves 
the  political  factor  out  of  account,  but  also  how  easily  the  latter  may 
acquire  an  importance  compared  with  which  that  of  any  decline  of  invest- 
ment opportunity  from  reasons  inherent  to  the  capitalist  process  would  be 
negligible,  even  if  it  did  occur  at  a  significant  rate  per  year. 

1  Even  if  it  had  been  still  more  important  than  it  was,  this  factor  would  have  to  be 
listed  with  a  qualification.  The  development  of  new  sources  of  power  and  their  competition 
with  the  old  ones  is,  of  course,  part  of  our  process  and  not  an  impediment  to  it.  On  princi- 
ple, it  is  immaterial  by  whom  the  new  sources  are  developed,  whether  by  public  or  private 
enterprise.  The  right  of  that  factor  to  a  place  in  the  list  of  spokes  in  the  economic  wheel 
rests  with  the  extent  to  which  the  federal  and  municipal  works  can  be  expected  to  be 
privileged  and  to  be  made  tools  of  attack  irrespectively  of  economic  (cost-accounting) 
rationality. 


1044  BUSINESS  CYCLES 

Second,  there  is  nothing  surprising  in  the  fact  that  under  the  circum- 
stances the  no  less  old  hostility  against  "monopoly  power'*  should  have 
asserted  itself  again  all  over  the  industrial  field.  But  "monopoly  "  really 
means  any  large-scale  business.  And  since  economic  "progress"  in  this 
country  is  largely  the  result  of  work  done  within  a  number  of  concerns 
at  no  time  much  greater  than  300  or  400,  any  serious  threat  to  the 
functioning  of  these  will  spread  paralysis  in  the  economic  organism  to  a 
much  greater  degree  than  a  similar  threat  to  the  corresponding  number  of 
concerns  would  in  any  other  country.  No  compensation  was  afforded  by 
the  federal  government's  extreme  anxiety  not  to  show  hostility  to  private 
business  in  general  or  to  do  anything  that  could  have  aroused  the  cry  of 
Government  in  Business,  because  the  contributions  of  the  favored  strata  to 
"progress"  and  their  investments  are  not  only  comparatively  small  but 
also,  to  a  large  extent,  induced  by  what  happens  in  the  world  of  big  busi- 
ness. That  hostility  propelled  or  facilitated  the  fiscal  and  labor  policies 
which  we  have  glanced  at  above.  Beyond  these  very  little  was  actually 
done;  but  much  was  foreshadowed  at  various  times,  even  before  the 
monopoly  investigation,  recently  instituted.  This  may  have  meant  noth- 
ing or  everything,  according  to  whether  or  not  the  threats — no  doubt, 
again  indefinite — were  taken  seriously  by  those  whose  decisions  they 
could  have  influenced.  But  it  should  be  observed  how  very  much  like 
"liquidity  preference  owing  to  vanishing  investment  opportunity"  the 
behavior  would  look  which  would  result  if  they  were. 

d.  It  will  be  seen  that  none  of  all  the  measures  mentioned  under 
our  three  headings  can,  if  considered  individually,  be  reasonably  held  to 
have  played  a  dominant  role  in  shaping  the  current  Juglar.  An  easy 
road  thus  seems  to  lead  toward  the  conclusion  adopted  as  a  matter  of  fact 
by  many,  if  not  most,  economists,  that  no  explanation  can  on  these 
lines  be  derived  for  the  lack  of  vitality  displayed  by  the  economic  process 
during  the  period  under  survey,  and  that  investment  opportunities  must, 
hence,  be  vanishing  from  causes  internal  to  that  process,  in  spite  of  all 
we  have  adduced  to  the  contrary.  It  is,  however,  suggested  that  the 
following  considerations  greatly  strengthen  the  case  for  the  adequacy  of 
that  explanation. 

First,  the  combined  effect  of  a  series  of  measures  unfavorable  to  invest- 
ment opportunity  can  evidently  not  be  evaluated  by  adding  up  the  effects 
which  each  of  them  would  have  had  in  the  absence  of  the  others.  We  have 
repeatedly  met  cases  of  this  kind.  For  instance,  the  reader  will  recall  that 
our  discussion  of  the  course  of  German  wage  rates  did  not  result  in  any- 
thing amounting  to  proof — or  rather  that  it  led  us  to  deny — that  this 
element  would  have  spelled  serious  disturbance  if  it  had  acted  alone;  but 
that,  since  it  did  not,  this  was  not  the  relevant  question  to  put.  Simi- 


THE  WORLD  CRISIS  AND  AFTER  1045 

larly,  we  might  in  the  case  before  us  make  even  larger  concessions  than  the 
writer  would  be  prepared  to  justify,  to  the  prevalent  tendency  to  under- 
estimate the  effects  of  any  or  all  the  individual  measures  we  have  glanced 
at,  and  nevertheless  have  to  conclude  that  their  combined  effects  were 
adequate  to  produce  the  observed  result.  The  individual  measures 
obviously  tended  to  reinforce  each  other.  "Objectively0 — i.e.,  irrespec- 
tively of  intentions  harbored  by  any  individuals1 — they  amounted  to 
systematic  attack  on  investment  opportunity  all  round:  it  was  frontally 
attacked  by  direct  reduction  of  revenues — or  the  operative  part  of  total 
net  revenues — through  taxation,  which  would  have  been  only  the  more 
effective  if  there  really  had  been  also  an  inherent  tendency  for  investment 
opportunity  to  shrink;  simultaneously,  it  was  attacked  in  the  rear  by 
increasing  costs;  and  both  attacks  were  supplemented  by  a  third — the 
attack  on  those  traditional  methods  of  management,  pricing,  and  financ- 
ing in  the  sphere  of  "big  business"  which  were  associated  with  the  latter's 
emergence  and  successes.  No  doubt,  opinions  still  may  in  all  fairness 
differ  as  to  the  importance  both  of  these  combined  attacks  and  of  the 
precise  points  in  the  industrial  structure  that  were  being  attacked.  But 
difference  of  opinion  is  not  possible  about  the  relevance  of  the  principle 
of  interpretation  which  the  writer  is  trying  to  stress.2 

Second,  the  mistake  involved  in  trying  to  arrive  at  an  estimate  of 
combined  effects  by  that  process  of  addition  is  not  more  serious  than  the 
mistake  of  confining  attention,  in  evaluating  either  isolated  or  combined 
effects,  to  the  wording  of  enactments,  congressional  declarations  of  policy, 
and  statements  of  the  chief  executive.  Economists  who  pride  themselves 
on  the  practical  bent  of  their  researches  could  really  be  expected  to  know 
that  the  personnel  and  methods  by  which  and  the  spirit  in  which  a  measure 
or  set  of  measures  is  administrated,  are  much  more  important  than  any- 
thing contained  in  any  enactment.  We  have  met  with  examples  above. 
The  events  surrounding  the  National  Labor  Relations  Act  will  again 
serve  to  illustrate  that  simple  truth,  particularly  if  we  compare  American 
with  English  experience  in  that  field :  it  should  be  obvious  that  in  the  one 

1  It  is  hoped  that  the  reader  will  not  object  to  the  above  statement  on  the  ground  that 
he  is  quite  sure  that  Mr.  X  or  Mr.  Y  does  not  harbor  any  hostile  intentions.     If,  however, 
this  objection  were  raised,  the  writer  would  not  reply  by  asking  how  the  reader  knows  that 
or  by  pointing  out  that  policies  are  hardly  ever  made  by  the  men  at  or  near  the  top  of  the 
official  ladder;  he  would  reply  that  intentions  are  entirely  irrelevant.     History  is  a  record  of 
"effects"  the  vast  majority  of  which  nobody  intended  to  produce. 

2  It  is  neglected  to  a  surprising  degree,  even  in  cases  in  which  there  is  no  aversion  to 
the   conclusion   implied.     Authorities   on   accounting  and   taxation   habitually   discuss 
effects  of  fiscal,  authorities  on  labor  discuss  labor  policies  in  isolation,  even  if  their  intention 
is  to  prove  the  presence  of  "injurious"  effects,  apparently  without  realizing  that  they 
thereby  reduce  their  arguments  to  exercises  in  general  theory. 


1046  BUSINESS  CYCLES 

case  effects  on  investment  opportunity  may  result  which  it  would  be 
absurd  to  expect  in  the  other.  This  already  covers  part  of  what  we  desig- 
nate by  the  term  Social  Atmosphere. 

But,  third,  this  atmosphere  should  also  be  listed  independently  as 
an  additional  factor  in  its  own  right.  It  is  surely  not  too  much  to  ask 
economists  to  realize  that  behavior  in  human  societies  differs  from  behav- 
ior in  animal  societies  or  in  physical  systems,  in  that  it  not  simply  reacts 
to  "disturbances"  but  to  interpretative  and  anticipative — correct  or 
false — diagnoses  of  them.  Real  or  supposed  drifts  and  trends  may  count 
as  much  as  or  more  than  facts,  threats  as  much  as  actions,  indefinite 
threats  more  than  specific  ones,  in  creating  the  psychic  environment  in 
which  the  nation's  work  has  to  be  done.  We  know  that  behind  those 
measures,  administrative  acts,  and  anticipations  there  is  something  much 
more  fundamental,  viz.,  an  attitude  hostile  to  the  industrial  bourgeoisie 
which  is  no  ephemeral  composite  of  individual  circumstances  and  political 
exigencies  of  the  day  but  the  product  of  the  same  social  process  that 
created  that  bourgeoisie.  Businessmen  presumably  do  not  hold  that 
theory.  But  they  need  not  hold  any  in  order  to  realize  that  there  is  in 
those  measures  and  programs  more  than  there  would  have  been  in  similar 
measures  and  programs  30  years  ago.  They  are  not  only,  but  they  feel 
threatened.  They  realize  that  they  are  on  trial  before  judges  who  have 
the  verdict  in  their  pocket  beforehand,  that  an  increasing  part  of  public 
opinion  is  impervious  to  their  point  of  view,  and  that  any  particular 
indictment  will,  if  successfully  met,  at  once  be  replaced  by  another. 
Again,  we  may  differ  in  our  estimates  of  the  importance  both  of  this  factor 
and  of  the  functions  it  tends  to  paralyze,1  but  it  should  not  be  overlooked. 

Fourth,  the  effects  of  all  that  on  investment  opportunity — if  the  reader 
prefer,  on  what  to  the  businessman  appears  as  an  investment  opportunity 
of  a  given  degree  of  attractiveness — were  greatly  enhanced  by  the  sud- 
denness of  the  change  of  scene.  In  order  to  convince  ourselves  of  this,  it 
is  only  necessary  to  reflect  that  any  major  change  in  the  relations  between 

1  Those  are,  of  course,  different  things.  Concerning  the  first,  the  writer  believes  that 
observations  are  available  which  point  to  the  conclusion  that  the  influence  of  "atmosphere" 
on  the  behavior  of  entrepreneurs  and  "capitalists"  was  considerable  but  admits  that  the 
nature  of  the  case  precludes  proof  and  that  any  attempt  at  proof  by  consequences  would  be 
circular.  Concerning  the  latter,  few  people  will  doubt  the  importance  of  those  effects  if 
the  argument  is  made  to  turn  on  aversion  to  investing.  But  difference  of  opinion  would  at 
once  appear  if  it  were  suggested  that  effects  of  the  hostile  atmosphere  can  also  consist  in 
impairing  the  efficiency  of  entrepreneurs  and  managers.  For,  as  pointed  out  in  Chap.  XIV, 
Sec.  C,  2,  it  is  part  of  the  creed  of  the  day  that  those  people  who  have  any  reason  to  feel 
threatened  do  not  fill  any  function  anyhow.  That  this  is  an  article  of  creed  and  not  based 
on  any  familiarity  with  the  facts  of  business  practice  any  teacher  of  economics  can  easily 
establish  for  himself:  the  belief  is  present  at  a  stage  in  life  in  which  nobody  can  have 
acquired  an  idea  of  what  it  means  to  run,  let  alone  create,  a  large-scale  concern. 


THE  WORLD  CRISIS  AND  AFTER  1047 

the  individual  and  the  state,  including  any  major  shift  in  favor  of  the 
latter  of  the  shares  in  total  private  revenue  earned,  involves  changes  in  the 
fundamental  habits  of  mind,  the  attitudes  to  life,  and  the  valuations  at 
least  of  those  who  are  immediately  concerned.  The  sociology  of  this 
need  not  detain  us.  But  as  a  matter  of  history  it  is  clear  that  such  changes 
usually  come  about  by  small  installments  and  as  the  result  of  a  slow 
process  of  education,  which  must  be  far  advanced  for  codification  of 
principles  into  a  new  body  of  law  to  be  a  success.  We  observe,  in  fact, 
that  the  modern  principles  of  English  taxation  took  about  30  years  to 
develop  and  "sink  in,"  and  that  the  beginning  of  the  modern  system  of 
English  social  policies  dates  back  to  at  least  the  eighties  of  the  nineteenth 
century,  when  the  ideas  of  Chamberlain  and  Dilke  spread  dismay  among 
their  colleagues  in  Gladstone's  second  administration.1  The  English 
bourgeoisie  was  thus  given  time  to  acclimatize.  This  is  why,  as  the  reader 
will  recollect,  we  have  not  in  the  English  case  used  the  factors  under  dis- 
cussion in  any  explanation  of  the  features  of  any  individual  Juglar  phase, 
or  of  any  short-run  phenomena  in  general,  but  only  in  our  explanation  of 
long-run  contours  and  levels. 

But  in  this  country  there  was  no  such  preparation;  hence,  there  was  a 
different  reaction.2  Barring  the  war  intermezzo,  there  was  nothing 
except  the  feeling  against  "monopolies"  and  utilities  to  indicate  any 
resentment,  and  that  was  of  the  middle-class  type  only  and  easy  to  keep 
in  hand.  On  the  whole,  the  businessman's  moral  world  was  the  nation's 
moral  world  right  up  to  the  crisis.  And  for  nearly  two  years  the  demo- 
cratic administration,  though  doing  many  things  which  were  felt  to  be 
"unorthodox"  by  its  friends  as  well  as  by  its  foes — measures  actually 
were,  as  we  have  seen,  not  so  very  much  out  of  keeping  with  American 
tradition — in  no  way  displayed  the  attitude  that  we  are  discussing  now 
but,  on  the  contrary,  signs  of  a  thoroughly  bourgeois  attitude.  The 
change  in  policy  dates  only  from  1934—1935.  It  therefore  followed  rather 
than  preceded  the  radicalization  of  the  public  mind,  which  in  conse- 
quence of  the  crisis  had  occurred  between  1930  and  1933  as  radicalization 
in  countries  in  which  authority  is  associated  with  military  values  will 
occur  in  consequence  of  military  defeat. 

The  analogy  with  the  German  breakdown  in  1918  suggested  by  the  last 
remark  indicates  the  line  on  which  we  should  explain  how  and  why  a 

1  It  would,  however,  be  more  just  to  date  from  the  preceding  Disraeli  administration 
and,  in  another  sense,  from  the  forties  (Lord  Ashley). 

2  Recognition  of  this  element  is  often  implied  in  current  talk  about  "lack  of  confidence," 
"strike  of  capital,"  and  a  "government-made  depression."     These  and  similar  phrases  are 
objectionable  on  several  grounds  and  indicative  of  naive  resentments  rather  than  of 
correct  diagnosis,  but  in  the  impression  they  are  intended  to  convey  there  is  an  element  of 
truth. 


1048  BUSINESS  CYCLES 

secular  process,  after  having  failed  to  assert  itself  to  any  practically  signif- 
icant extent  for  fully  forty  years  after  the  closing  of  the  frontier,  then 
suddenly  became  the  dominant  factor  of  the  political  situation.  In 
doing  so,  we  should  no  doubt  have  to  go  into  many  circumstances  peculiar 
to  the  American  environment  in  general  and  to  American  politics  in 
particular,  in  order  to  understand  the  details  of  the  change  in  attitude  and 
of  the  resulting  political  pattern.  But  the  fact,  the  broad  causes,  and  the 
effects  on  business  behavior  are  sufficiently  obvious  to  establish  our  point 
without  any  analysis  of  details.1  There  are,  however,  two  aspects  which 
cannot  be  passed  by. 

On  the  one  hand,  we  have  insisted  above  on  the  importance  of  person- 
nel and  of  methods  of  administration.  New  measures  as  well  as  new 
attitudes  must  be  implemented  by  a  skilled  civil  service.  In  any  case 
they  set  a  difficult  task  to  even  the  most  experienced  bureaucracy.  As 
a  rule,  however,  reforming  governments  enjoy  at  least  the  advantage  of 
having  that  indispensable  tool  ready  at  hand — in  most  historical  instances 
it  grew  up  along  with  the  tendencies  which  they  represent.  This 
happened,  for  example,  in  England,  while  in  Germany  the  regime  of  1918 
was  able  to  take  over  from  its  predecessor  both  an  excellent  civil  service 
and  a  state-broken  public.  In  this  country  .a  new  bureaucracy  had 
suddenly  to  be  created.  However  good  part  of  the  material  on  which  it 
was  necessary  to  draw,  and  however  creditable,  considering  the  circum- 
stances, the  performance  of  a  great  many  individuals  and  groups  may 
have  been,  there  was  no  experience,  no  esprit  de  corps,  no  clear  idea  even  of 
what  civil  service  is  and  what  it  can  and  cannot  do.  No  less  inexperi- 
enced— to  the  point  of  not  seeing  the  fundamental  administrative 
problems  at  all — were  the  men  in  whose  hands  that  unwieldly  apparatus 
was  put.  The  tact,  the  reserve,  the  savoir-faire  which  are  second  nature 
to  a  seasoned  bureaucracy  were  alike  absent.  Enthusiastic  individuals 
and  groups  developed  their  own  policies  and  tried  to  push  them  with 
Congress  and  the  public,  scornfully  refusing  counsels  of  self-denial  and 
patience.  In  consequence,  that  sense  of  indefinite  threat  was  immeasur- 
ably increased.  English  policies  may  be  felt  to  be  equally  or  more 
oppressive,  but  they  are  never  aggressive:  spectacular  manifestations  of 

1Some  of  the  economic  causes  of  the  sudden  change  of  scene  have,  however,  been 
pointed  out  previously,  among  them  the  agrarian  situation,  the  last  epidemic  among 
banks,  the  stock  exchange  crash,  unemployment.  The  main  factors  which  are  and  must 
remain  absent  from  our  sketch  are  the  structure  and  technique  of  American  politics  and 
the  role  played  by  the  "intellectuals."  A  complete  analysis  would  also  have  to  take 
account  of  random  configurations,  of  which  one  has  been  mentioned,  viz.,  the  fact  that  a 
presidential  election  occurred  a  few  months  after  the  lower  turning  point,  exactly  at 
the  time  at  which  the  sociopsychological  "hangover"  of  the  depression  should  be  at  its 


THE  WORLD  CRISIS  AND  AFTER  1049 

aggressiveness  proceed  only  from  quarters  that  are — by  all  parties1 — 
firmly  held  in  check,  and  never  from  members  of  the  public  service.  The 
methods  of  the  latter  may  be  likened  to  deerstalking  and  tend  to  minimize 
trouble  and  disturbance  caused  by  any  given  measure.  Administrative 
methods  in  this  country  tend  to  maximize  them  and  are  more  like  those 
of  the  fox  hunt — and  this  makes  a  lot  of  difference. 

On  the  other  hand,  sudden  change,  unless  of  the  Russian  type,  is  of 
necessity  imperfect  change.  It  impinges  upon  a  set  of  economic  and  polit- 
ical conditions  which  are  very  unequally  ripe.  This  puts  advocates  and 
opponents  of  new  departures  in  false  positions,  adulterates  arguments, 
and  makes  it  impossible  to  face  issues  squarely.  In  England  the  question 
of  the  employment  of  nonunionized  labor,  for  instance,  having  been 
allowed  to  mature,  is  now  one  of  secondary  importance.  In  this  country 
it  cannot  even  be  frankly  put,  yet  it  is  at  the  bottom  of  much  strategy 
and  struggle  which,  precisely  because  the  issue  is  not  ripe  for  decision, 
must  be  expected  to  remain  for  some  time  a  source  of  difficulties  and  losses 
to  all  parties  concerned.  But  the  standard  instance  is  the  policy  followed 
with  respect  to  public  utilities.  Here,  if  anywhere,  there  was  an  all  but 
united  public  opinion,  united  at  least  in  its  hostility  to  the  private  interests 
involved.  Moreover,  European  experience  suggestively  pointed  toward 
nationalization  of  power  production  and  transmission,  which  could  have 
been  carried  without  any  shock  to  "business  confidence "  if  the  interests  of 
investors  had  been  fully  safeguarded,  and  with  but  a  sharp  and  short  one 
if  they  had  been  sacrificed — always  provided  that  there  were  no  clenched 
fists  or  indeterminate  threats  of  other  nationalizations  to  follow.  Yet 
it  was  not  even  attempted.2  The  clenched  fists  and  indeterminate  threats 
were  all  the  more  in  evidence,  however,  and  the  result  was,  as  we  have  seen, 
to  paralyze  one  force  without  substituting  another.  This  will  always 
result  from  raising  issues  before  they  can  be  effectively  dealt  with  and 
illustrates  what  above  has  been  described  as  deadlock.  To  deny 
that  this  impairs  the  efficiency  of  the  economic  engine  or,  if  we  retain 
the  slogan,  reduces  investment  opportunity,  would  seem  to  the  writer 
unreasonable. 

1  The  official  leaders  of  the  Labor  party,  hence,  frequently  find  themselves  described 
as  what  in  plain  English  would  have  to  be  called  dunces  and  traitors.     But  it  does  not  mat- 
ter, and  everyone  knows  that  action  upon  the  views  of  radical  wings  would  speedily  end  in 
discomfiture. 

2  A  move  toward  encouraging  municipalities  to  acquire  the  works  and  distributive 
systems  that  serve  them  has  recently  been  made.    It  only  illustrates  our  point,  both  by  its 
weakness  and  by  its  timing.     Another  illustration  is  afforded  by  the  practice  of  the  PWA 
of  facilitating  the  construction  of  distribution  systems  duplicating  the  privately  owned 
ones  in  the  sector  of  the  TVA  by  charging  less  than  cost  and  by  lending  the  remainder  at 
low  rates  to  municipalities. 


1050  BUSINESS  CYCLES 

If  the  above  considerations  are  given  their  proper  weight,  there  should 
not  be  left  much  doubt  as  to  the  adequacy  of  the  factors  external  to  our 
process1  to  account  both  for  the  disappointing  features  in  the  current 
Juglar  and  for  the  weakness  of  the  response  of  the  system  to  government 
expenditure,  in  particular  for  the  failure  of  the  latter  to  affect  investment 
and  employment  more  than  it  did.  It  cannot  be  proved  in  the  sense  in 
which  a  mathematical  theorem  can,  that  the  balloon  shriveled,  not  from 
causes  inherent  to  its  structure,  but  because  the  air  was  being  sucked  out 
of  it.  It  is,  however,  highly  plausible  and,  after  all,  only  what,  if  we  clear 
our  minds  of  cant,  we  should  expect  to  occur  in  transitional  stages. 
Prognosis  would,  in  this  country  more  than  in  any  other,  have  to  take 
account  of  the  likelihood  that  there  will  be  intermissions  or  even  reversals ; 
of  the  effects  of  "acclimatization";  and  of  the  fact  that,  if  our  schema  is 
to  be  trusted,  recovery  and  prosperity  phases  should  be  more,  and 
recession  and  depression  phases  less  strongly  marked  during  the  next 
three  decades  than  they  have  been  in  the  last  two.  But  the  sociological 
drift  cannot  be  expected  to  change. 

1  Let  us  add,  however,  for  the  last  time,  "in  the  narrow  sense  adopted  for  the  purposes 
of  this  book."  In  a  wider  sense  those  factors  and  the  mentality  or  moral  code  behind  them 
are  not  external  to  the  process  of  economic  evolution  but  part  of  it,  a  part  as  essential  and 
unavoidable  as  any  other  and,  in  particular,  as  any  "objective"  shrinkage  of  investment 
opportunity  could  be.  The  above  argument  would,  hence,  be  completely  misunderstood 
if  it  were  taken  to  imply  that,  "only"  politics  or  humors  being  the  matter,  pristine  vigor  of 
the  capitalist  process  could  easily  be  restored  at  the  next  swing  of  the  electoral  pendulum. 
As  far  as  that  goes,  the  practical  implications  of  our  diagnosis  do  not  differ  much  from  those 
of  the  theory  of  vanishing  investment  opportunity  in  its  usual  acceptance.  Even  govern- 
ment spending  as  a  permanent  policy  could  be  rationally  defended  on  our  diagnosis:  the 
pattern  resulting  from  the  action  of  inhibiting  factors  would  in  all  respects  be  similar  to  the 
pattern  envisaged  by  the  saving-investment  theory;  it  would  display  the  same  lack  of 
resilience  and  the  same  tendency  toward  subnormal  quasi-equilibria;  in  particular,  it  would 
always  produce  or  reproduce  extensive  unemployment.  Therefore,  government  spending 
would,  given  the  general  will  to  conserve  those  inhibiting  factors,  always  suggest  itself 
as  a  remedy  for  shortrun  difficulties  each  application  of  which  would  impose,  under  penalty 
of  breakdown,  the  application  of  the  next  dose.  Fear  of  such  breakdowns  may  in  the  end 
become  the  dominant  motive  even  among  those  who  on  principle  are  most  strongly  opposed 
to  spending  policies. 


APPENDIX 

Description  of  the  Statistical  Material  Embodied  in  the  Charts 
CHART  I.    ARITHMETIC  SCALE 

Three  sine  curves  with  periods  respectively  of  684,  114,  and  88  months  (or  57,  9j£,  and 
3;Hi  years)  and  with  amplitudes  roughly  proportional  to  the  periods,  i.e.,  in  the  relation  of 
18,  3,  and  1,  have  been  plotted  separately.  The  fourth  curve  presents  their  sum. 


<.>-»-©)• 


(3)  T  =  si 

fA\  1Q      •       /360A°_LQ      '       ^860A°_L      •        /360\° 

(4)  y  =  18  sm  (—tj    +  3  sm  {—t)    +  sm  (— t) 

CHART  II.     ARITHMETIC  SCALE 

The  graph  represents  the  derivative  (time  rate  of  change)  of  the  sum  of  three  sine 
curves  with  periods  of  684,  114,  and  88  months. 

CHART  III.     ARITHMETIC  SCALE 

(1)  Average  Railroad  Revenue  Freight  Loaded.     United  States.     Weekly  figures. 

(2)  Four-item  Moving  Average  of  Three-item  Moving  Average  of  Two-item  Moving 
Average  of  Series  (1). 

(3)  Superseasonal  Normal.     This  is  a  freehand  curve  drawn  through  the  inflection 
points  of  the  moving  average  curves. 

(4)  Deviations  of  Series  (2)  from  Series  (8)  multiplied  by  the  "Inflation  Factor." 

For  a  discussion  of  the  Frisch  method  in  general  the  reader  is  referred  to  Chap  V.  For 
a  discussion  of  the  Inflation  Factor  cf.  Horst  Mendershausen,  Annual  Survey  of  Statistical 
Technique:  Methods  of  Eliminating  Changing  Seasonal  Fluctuations,  Eeonometrica,  vol.  V, 
no.  3,  July  1937,  p.  244. 

Let 

/  =  Inflation  Factor 
Then 

_  sin  2X      sin  3X      sin  4X 
2  sin  X    8  sin  X    4  sin  X 

where 

»-..? 

and 

p  =  twice  the  time  distance  between  normal  points 
and 

D  =  1,  the  time  unit  in  which  p  is  measured. 
1051 


1052  BUSINESS  CYCLES 


CHART  IV.    RATE  OF  PERCENTAGE  CHANGE  OF  PRICES.    ARITHMETIC 

SCALE 

The  rate  of  percentage  change  for  each  series  is  computed  by  applying  to  the  logarithms 
of  the  indices  the  formula 

Dk  -  ^(iA*  -f  iAfc+0  -  K2(sAHi 
for  annual  indices  or  the  formula 

Dk  «  2(iA*  +  iA*+i)  ~  MGAjfe+i 

% 

for  quarterly  indices.     The  antilogs  of  the  D  items  minus  100  %  give  the  rates  of  percentage 
change. 

The  rate-of -change  formulae  are  applied  to  logs  rather  than  absolute  figures  for  reasons 
analogous  to  those  leading  to  the  use  of  logarithmic  rather  than  arithmetic  scales  on  charts. 
The  importance  of  a  change  in  the  series  is,  hence,  made  to  depend  on  the  relative  (per- 
centage or  geometric)  movement. 

(1)  United  States.     Wholesale  Prices.     1842-1913.     The  data  have  been  taken  by 
permission  from  Prices  by  G.  F.  Warren  and  F.  A.  Pearson,  published  by  John  Wiley  & 
Sons,  Inc.,  New  York,  pp.  12-13  (see  also  the  discussion  in  the  text  and  in  the  note  on  p.  14), 
and  built  into  quarterly  12-month  averages  of  the  monthly  items. 

(2)  Germany.     Wholesale  Prices.     1879-1913.     For  the  years  from  1879  to  1902  it 
has  seemed  best  to  use  the  monthly  index  compiled  by  Otto  Schmitz,  Die  Bewegung  der 
Waarenpreise  in  Deutschland  von  1851  bis  1902,  Berlin,  1903.     This  index  is  an  unweighted 
arithmetic  average  of  the  prices,  published  since  1879  by  the  Imperial  Statistical  Office,  of 
29  articles,  raw  materials  almost  entirely.     From  1898  to  1913  the  author  has,  by  permis- 
sion, used  the  monthly  index  (average  1898-1913  =  100)  of  10  commodities  (cotton  yarn, 
cotton  cloth,  raw  silk,  wool  (2),  calf  skins,  pig  iron  (2),  lead,  tin)  constructed  by  E.  W.  Axe 
and  H.  M.  Flinn,  An  Index  of  Business  Conditions  for  Germany,  1898-1914,  Review  of 
Economic  Statistics,  vol.  VII,  1925,  pp.  282-285.     Quarterly  12-month  averages  of  the  two 
indices  were  spliced  together,  bases  being  shifted  so  as  to  make  them  coincide  as  nearly  as 
possible  from  1898  to  1902.     At  the  time  this  seemed  the  best  thing  to  do.     In  his  later 
work  the  writer  availed  himself  of  the  new  and  incomparably  more  satisfactory  index  of 
the  Institut  fiir  Konjunkturforschung,  leaving  however  the  chart  as  it  was. 

(3)  United  Kingdom.     Wholesale  Prices.     1850-1913.     From  1850  to  1870  the  annual 
Sauerbeck  index  (Prices  of  Commodities  and  the  Precious  Metals,  Journal  of  the  Royal 
Statistical  Society,  vol.  XLIX,  September  1886,  p.  648)  was  used  with  the  permission  of  the 
Royal  Statistical  Society.     This  index  is  a  simple  average  (1867-1877  =  100)  of  56  whole- 
sale prices,  almost  entirely  of  raw  materials:  22  foods,  8  minerals,  11  textiles,  and  15  other 
materials.     Cotton  and  other  articles  are  counted  more  than  once.     From  1870  to  1885 
recourse  has  been  had  to  the  annual  Board  of  Trade  Index  (see  Cmd.  6955,  1913,  p.  308) 
which  is  a  weighted  aggregate  (1900  =  100)  of  the  prices  of  foods  and  raw  materials,  the 
articles  being  weighted  according  to  the  estimated  consumption  during  the  years  1881  to 
1890.     From  1885  to  1913  quarterly  12-month  averages  were  computed  from  the  monthly 
index  of  the  editor  of  the  Statist  (cf.  Wholesale  Prices  of  Commodities  in  1921,  Journal  of 
the  Royal  Statistical  Society,  vol.  LXXXV,  March  1922,  p.  275),  a  continuation  of  Sauer- 
beck's index.     The  three  indices  were  spliced  together  graphically  on  the  basis  of  over- 
lapping years. 

CHART  V.     BRITISH  PREWAR  PULSE.    LOG  SCALE 

(1)  Interest  Rate.  1824-1913.  From  1824  to  1850  use  was  made  by  permission  of 
the  series  of  quarterly  average  discount  rates  in  London  (best  bills)  as  published  by  Norman 


APPENDIX  1053 

J.  Silberling  in  British  Prices  and  Business  Cycles,  Review  of  Economic  Statistics,  Supple- 
ment, vol.  V,  1923,  p.  257  (see  footnote  2,  p.  242).  For  the  most  part  these  are  the  rates 
which  were  charged  by  the  firm  of  Overend,  Gurney  &  Co.  The  data  were  built  into  four- 
quarter  moving  totals.  From  1850  to  1857  quarterly  12-month  moving  averages  of  the 
monthly  average  rates  charged  on  best  bills  by  Overend,  Gurney  &  Co.  have  been  computed 
from  the  figures  in  Parliamentary  Papers,  1857,  vol.  X,  Part  I,  p.  464.  From  1858  to  1861 
the  quarterly  average  market  rates  on  high-class  bills  were  taken  from  R.  H.  I.  Palgrave, 
Bank  Rate  and  the  Money  Market,  p.  83,  by  permission  of  the  publisher,  John  Murray. 
From  1862  to  1913  the  curve  has  been  based  on  the  figures  of  the  Lombard  Street  rate,  at 
the  beginning  of  each  month,  on  two  or  three  months'  bills  as  published  in  the  London 
Economist's  Annual  Commercial  History  and  Review.  The  four  series  were  spliced  by 
means  of  overlapping  periods. 

(2)  Production.     1785-1913.    This   is   Dr.   W.   Hoffmann's   annual   index   of  total 
industrial  production,  cf.  Ein  Index  der  industriellen  Produktion  ftlr  Grossbritannien  seit 
dem  18.     Jahrhundert,  Weltwirischaftliches  Archiv.,  vol.  XL,  1934  II,  pp.  396-398,  the  use 
of  which  has  been  kindly  permitted  by  the  author  and  the  editor,  Professor  Predohl.     In 
1785,  10  commodities  are  represented.     See  that  article  for  commodities  and  weights. 
Most  of  the  50  series  finally  included  begin  about  1800.     It  is  believed  that  from  that  year 
on  approximately  two-thirds  of  English  industry  is  covered.     For  purposes  of  weighting, 
the  index  is  broken  up  into  six  shorter  periods  and  also  into  broad  industrial  groups. 
Indices  are  separately  computed  for  producers'  goods  and  consumers'  goods.     Foods, 
textiles,  materials,  chemicals,  iron  and  steel  and  machine  industries,  mining,  transporta- 
tion, and  electricity  are  the  main  divisions  covered.     The  reader's  attention  is  invited  to 
Dr.    Hoffmann's   recent  book  on   the   subject,   Wachstum   und   Wachstumsformen   der 
Englischen  Industriewirtschaft. 

(3)  Prices  at  Wholesale.     1785-1913.    From  1785  to  1850  Mr.  Norman  J.  Sterling's 
quarterly  index  (Review  of  Economic  Statisticst  Supplement,  quoted  above)  has  been  used 
by  permission.     It  is  a  simple  geometric  mean  of  the  prices  of  35  commodities,  mostly  raw 
materials  (1790  =  100).     Four-quarter  moving  totals  were  plotted.     From  1850  on,  the 
series  are  as  described  under  Chart  IV  (3).     All  splicing  was  done  on  the  basis  of  over- 
lapping years. 

(4)  Deposits  plus  Circulation.     1880-1913:  deposits  in  joint  stock  and  private  banks 
in  Great  Britain  (exclusive  of  the  Bank  of  England).     The  circulation  is  total  note  circula- 
tion in  Great  Britain  at  the  beginning  and  middle  of  each  year.     Both  series  were  taken 
from  the  Economist's  Reports  of  Joint  Stock  Banks  of  the  United  Kingdom. 

CHART  VI.    UNITED  STATES  PREWAR  PULSE.    LOG  SCALE 

(1)  Deposits  plus  Circulation.     1870-1913.     The  data  are  call-date  figures,  five  each 
year,  for  "individual  deposits"  and  "notes  outstanding"  of  all  national  banks  outside 
New  York  City,  taken  by  permission  of  the  Review  of  Economic  Statistics  from  A.  A.  Young, 
An  Analysis  of  Bank  Statistics  for  the  United  States,  Harvard  University  Press,  Cam- 
bridge, Mass.,  1928,  pp.  8-13.     A  five-item  moving  total  centered  on  the  third  item  has 
been  plotted.     "Individual  deposits"  are  individual  deposits  as  reported  by  the  Comptrol- 
ler of  the  Currency  minus  clearing-house  exchanges.     They  do  not  include  government  or 
interbank  deposits.     Before  Apr.  26,  1900,  however,  some  banks  counted  deposits  due  to 
savings  banks  as  individual  deposits.     "Notes  outstanding,"  of  course,  do  not,  except  very 
imperfectly,  indicate  variations  in  national  bank  notes  in  circulation. 

(2)  Interest  Rate.     Commercial  paper  rate.     1831-1913.     The  monthly  data  were 
taken,  with  the  permission  of  Colonel  Leonard  P.  Ayres,  from  his  chart  Business  Activity 
and  Four  Price  Series,  Cleveland  Trust  Company  Bulletin  for  July  1932.     The  sources 
quoted  by  Colonel  Ayres  are:  from  1831  to  1860,  E.  B.  Bigelow,  The  Tariff  Question,  Boston, 


1054  BUSINESS  CYCLES 

1862;  from  1861  to  1865,  J.  G.  Martin,  Seventy-three  Years'  History  of  the  Boston  Stock 
Market,  Boston,  1871;  and  from  1866  to  1913,  W.  L.  Crum,  Cycles  of  Rates  in  Commercial 
Paper,  Review  of  Economic  Statistics,  vol.  V,  January  1928,  p.  28.  Quarterly  12-month 
totals  of  the  monthly  items  were  plotted. 

(8)  Output  of  Manufacturing  and  Mining.  1868-1918.  The  curve  is  plotted  on 
annual  figures  taken  by  permission  from  Professor  W.  M.  Persons,  Forecasting  Business 
Cycles,  published  by  John  Wiley  &  Sons,  Inc.,  New  York.  This  index  is  a  combination  of 
two  indices  given  separately;  both  are  weighted  arithmetic  averages,  with  1909-1913  =  100 
as  base,  manufacturing  being  given  a  weight  of  %  in  the  combination.  See  W.  M.  Persons, 
op.  cit.,  pp.  172-173,  for  a  description  of  the  index  and  its  two  components. 

(4)  Wholesale   Prices.     1831-1913.     Monthly.    See   Chart   IV    (1).     Quarterly    12- 
month  averages  were  plotted. 

(5)  Crop    Production.     1870-1913.     Professor    W.    M.    Persons'    annual  x  index,    a 
weighted  arithmetic  average  (1909-1913  =  100).     For  description  see  op.  cit.,  pp.  171- 
172.     Acknowledgements  are  again  due  to  John  Wiley  &  Sons,  Inc.,  New  York. 

CHART  VII.     GERMAN  PREWAR  PULSE.    LOG  SCALE 

(1)  Wholesale    Prices.     1879-1913.     Monthly.     See    Chart   IV    (2).     Quarterly    12- 
month  averages. 

(2)  Industrial     Production     (Manufacturing     and     Mining).     1860-1913.     Annual. 
The  index  of  the  Institut  fur  Konjunkturforschung,  see  Rolf  Wagenfiihr,  Die  Industrie- 
wirtschaft,   Vierteljahrshefte  zur  Konjunkturforschung,  Sonderheft  31,   1933,  p.  58.     The 
index  is  computed  from  57  series  weighted  according  to  the  number  of  employees  and 
installed  horsepower  in  1907.     The  author's  thanks  are  due  to  the  President  of  the  Institut, 
Professor  Wagemann,  for  permission  to  use  it. 

(8)  Berlin  Market  Discount  Rate.  1868-1913.  From  1868  to  1898  the  monthly  data 
used  are  those  published  by  the  London  Economist's  Annual  Commercial  History  and 
Review.  From  1899  to  1913  monthly  figures  were  taken  from  E.  W.  Axe  and  H.  M.  Flinn, 
see  quotation  under  Chart  IV  (2).  They  are  averages  of  the  high  and  low  for  the  first 
week  of  each  month.  Quarterly  12-month  averages  were  computed. 

(4)  Note  Circulation  and  Credit  Accounts  of  Joint  Stock  Banks.  1884-1913.  Annual. 
Kreditoren  plus  Notennmlauf  for  the  end  of  each  year  from  Der  Deutsche  Oekonomist,  vol. 
XXIV,  1906,  pp.  464  and  466;  and  vol.  XXXII,  1914,  pp.  538  and  580. 

CHART  VIII.     NINE- YEAR  MOVING   AVERAGES  OF  PRICES.    LOG  SCALE 

The  following  series  were  put  in  the  form  of  annual  averages  from  which  were  computed 
9-year  moving  averages  centered  on  the  fifth  year. 

(1)  United  Kingdom.     Wholesale  Prices.     1779-1913.     For  description  of   material 
see  above  under  Chart  V  (3). 

(2)  Germany.     Wholesale  Prices.     1851-1913.     From  1851  to  1902  the  index  is  that 
of  Otto  Schmitz,  see  Chart  IV  (2),  which  until  1879,  when  the  index  of  the  Imperial  Sta- 
tistical Office  begins,  is  an  unweighted  average  of  the  Hamburg  prices  of  24  raw  materials. 
For  1903-1912  figures  were  taken  from  the  British  Board  of  Trade  Enquiry  into  Working- 
class  Rents  and  Retail  Prices,  etc.,  1912,  Cmd.  6955,  1913,  p.  356,  a  continuation  of 
Schmitz's  index. 

(3)  United  States.     Wholesale  Prices.     1797-1913.     This  is  the  series  described  in 
Chart  IV  (1). 

CHART  IX.    ARITHMETIC  SCALE 

The  original  data  (dots)  are  annual  wholesale  prices  in  the  United  States,  1790-1920, 
from  the  U.  8.  Department  of  Agriculture  Bulletin  999,  p.  2,  the  5-year  average,  1909-1914, 
being  equal  to  100. 


APPENDIX  1055 

(1)  A  3-year  on  a  2-year  moving  average  (graphically  determined). 

(2)  Curve  through  inflection  points  of  10-year  cycle. 
(8)  Curve  through  inflection  points  of  22-year  cycle. 
(4)  Curve  through  inflection  points  of  50-year  cycle. 

CHART  X.    PERCENTAGE  DEVIATIONS  OF  PRICES  FROM  9-YEAR 
MOVING  AVERAGE.    ARITHMETIC  SCALE 

The  annual  averages  of  the  series  used  for  Chart  VIII  were  divided  by  the  corresponding 
items  of  their  9-year  moving  averages  and  expressed  as  percentages  minus  100. 

CHART  XI.     UNITED   STATES  PRICES.     1840-1913.     ANNUAL.    LOG  SCALE 

(1)  Consumers'  Goods.     Wholesale.     From  1840  to  1890  the  index  (1890  =  100)  was 
computed  by  M.  J.  Fields  from  the  data  contained  in  the  Aldrich  report  of  1893  (Committee 
of  Finance,  United  States  Senate,  Fifty-second  Congress,  Second  Session,  Report  1394, 
Part  I,  pp.  91-94).     It  is  a  simple  arithmetic  average  of  the  three  groups:  (1)  food,  (2) 
clothing,  and  (3)  house  furnishings.     From  1890  the  index  of  the  U.  S.  Bureau  of  Labor 
Statistics  Bulletin  284,  pp.  48-49,  has  been  used.     See  also  Bulletin  149.     The  two  indices 
were  spliced  together. 

(2)  Foods.     Wholesale.     The  curve  is  based  on  figures  taken  by  permission  from  G.  F. 
Warren  and  F.  A.  Pearson,  Prices,  published  by  John  Wiley  &  Sons,  Inc.,  New  York, 
pp.  25-27.     1910-1914  =  100. 

(3)  Household  Furnishings.     Wholesale.     Same  source  as  (2). 

(4)  Textile  Products.     Wholesale.     Same  source  as  (2). 

(5)  Producers'  Goods.     Wholesale.     From  1840  to  1890  the  index  (1860  =  100)  was 
computed  by  M.  J.  Fields  from  data  in  the  Aldrich  Report  (see  Series  (1)).     It  is  a  simple 
arithmetic  average  of  two  groups,  metals  and  implements  (excluding  pocket  knives)  and 
lumber  and  building  materials.     From  1890  to  1913  the  index  is  the  one  published  by  the 
Bureau  of  Labor  Statistics  Bulletin  284  [see  Series  (1)].     The  two  indices  have  been  spliced. 

(6)  Metals  and  Metal  Products.     Wholesale.     Same  source  as  (2). 

(7)  Basic  Commodities.     Wholesale.     Same  source  as  (2).     This  index  includes  the 
prices  of  30  basic  commodities,  farm  products,  minerals,  textiles,  and  the  like. 

CHART  XII.    PIG  IRON  CONSUMPTION.     ANNUAL.    LOG  SCALE 

(1)  United  Kingdom.     1854-1913.     The  original  sources  are  the  Annual  Reports  on 
Mines  and  Quarries,  issued  by  the  Home  Office  and  the  Annual  Statements  of  the  Trade 
of  the  United  Kingdom.     The  series  was  taken  from  Parliamentary  Papers,  Cmd.  2145, 
pp.  24-25,  and  the  Seventeenth  Abstract  of  Labour  Statistics,  p.  44.     It  is  also  given  by 
A.  C.  Pigou,  Industrial  Fluctuations,  2d  ed.,  London,  1929,  p.  386. 

(2)  United  States.     1855-1913.     From  1855  to  1870  this  series  was  constructed  by 
E.  M.  Hoover  from  production  figures  of  the  American  Iron  and  Steel  Institute  and  imports 
for  calendar  years.     These  import  figures  were  obtained  by  interpolation  between  fiscal- 
year  import  totals  as  given  in  the  1871  Annual  Report  of  the  Institute.     Exports  and 
changes  in  stocks  were  ignored;  but  as  late  as  1870-1871  the  total  United  States  exports  of 
pig  iron  were  only  3,480  gross  tons  or  0.02  per  cent  of  the  1870  production.     From  1871  to 
1913  the  series  of  estimates  of  the  American  Iron  and  Steel  Institute,  published  in  its  Annual 
Statistical  Reports  and  based  on  data  for  production,  imports,  exports  and  (since  1874)  net 
change  in  stocks  held  was  used  with  its  permission.     Data  for  1871  to  1873,  being  given  in 
net  tons,  were  converted  to  gross  tons. 

(3)  Germany.     1860-1913.     The  series  (also  given  by  S.  Kuznets,  Secular  Movements 
in  Production  and  Prices,  pp.  469-470)  is  from  the  Statistisches  Jahrbuch  ftir  das  Deutsche 


1056  BUSINESS  CYCLES 

Reich.    The  figures  have  been  converted  from  metric  tons  (2,205  Ib.)  to  gross  tons  (2,240 


CHART  XIII.    RATE  OF  PERCENTAGE  CHANGE  OF  PIG  IRON 
CONSUMPTION.    ANNUAL.    ARITHMETIC  SCALE: 

The  rates  of  percentage  change  of  the  three  series  plotted  on  Chart  XII  were  calculated 
by  the  use  of  the  formula  explained  in  the  description  of  Chart  IV. 

CHART  XIV.    UNITED  STATES.    LOG  SCALE 

(1)  Cotton  Production.     1840-1911.     The  figures  (for  the  "cotton  year")  were  taken 
from  Circular  32,  U.  S.  Department  of  Agriculture,  The  Cotton  Crop  of  the  United  States, 
1790-1911.    Linters  are  included  from  1899  on. 

(2)  Railway  Freight  Ton-miles.     1852-1913.     Annual.     From  1852  to  1887  the  series 
was  taken  by  permission  from  Mr.  Carl  Snyder's  Business  Cycles  and  Business  Measure- 
ments, p.  238  (see  also  note,  p.  39),  published  by  The  Macmillan  Company,  New  York, 
1927.     This  series  gives  net  ton-miles  of  freight  carried  on  Class  I  railroads  estimated  from 
the  figures  of  the  principal  lines  as  reported  in  Poor's  Manual.     From  1888  to  1913  the 
series  was  taken  from  Railway  Statistics  of  the  United  States  of  America,  1916,  p.  99, 
prepared  by  S.  Thompson,  Bureau  of  Railway  News  and  Statistics,  Chicago,  1917.     Switch- 
ing and  terminal  companies  are  included  to  1908.     These  data  refer  to  Class  I,  Class  II, 
and  Class  III  railroads  and  fiscal  years. 

(3)  Coal  Production  (Anthracite  and  Bituminous).     1839-1913.     Annual.     This  index 
(1926-1930  =  100)  was  taken  by  permission  from  G.  F.  Warren  and  F.  A.  Pearson,  The 
Physical  Volume  of  Production  in  the  United  States,  Cornell  University  Agricultural  Experi- 
ment Station  Memoir  144,  Ithaca,  1932. 

(4)  Cotton  Consumption.     1862-1913.     Annual.     The  figures  are  from  the  Statistical 
Abstract  and  refer  to  "cotton  years."     (See  Census  Bureau  Bulletin  166,  p.  57). 

(5)  Building  Permits  Index.     1874-1913.     Annual.     This  series  (also  used  by  A.  F. 
Burns,  Production  Trends  in  the  United  States  since  1870,  pp.  302-303)  has  been  compiled 
by  Mr.  Carl  Snyder  and  published  (from  1882  on)  in  his  book  on  Business  Cycles  and 
Business  Measurements,  p.  275.     The  writer's  acknowledgments  of  permission  to  use  it 
are  due  to  Mr.  Snyder  and  The  MacMillan  Company.     It  represents  an  index  of  building 
permits  in  from  one  to  seven  cities  (seven  since  1895)  divided  by  an  index  of  changes  in 
costs  of  construction.     The  item  for  1879  (not  given  by  Mr.  Snyder)  was  interpolated  by 
referring  to  the  behavior  of  pig  iron  production  which  closely  paralleled  the  deflated 
building-permits  index  at  that  time.     The  assumption  was  that  the  ratio  of  1872f  879  *° 
187^{  880  was  the  same  in  the  two  series. 

(6)  Interest  Rate.     See  Chart  VI  (2). 

CHART  XV.    INDUSTRIAL  PRODUCTION.    ANNUAL.    LOG  SCALE 

(1)  United  States.     See  Chart  VI  (3). 

(2)  Germany.     See  Chart  VII  (2). 
(S)  Great  Britain.    See  Chart  V  (2). 

CHART  XVI.    RATE  OF  PERCENTAGE  CHANGE  OF  INDUSTRIAL 
PRODUCTION.    ANNUAL.    ARITHMETIC  SCALE 

See  Chart  XV  for  the  series  used  and  Chart  IV  for  the  formula. 
CHART  XVII.    PRODUCTION  (AND  CONSUMPTION).    ANNUAL.    LOG  SCALE 

(1)  United  States.     Producers'  Goods.     1870-1913.     This  index  of  the  production  of 
capital  equipment  has  been  constructed  by  E.  M.  Hoover.    1890-1900  =  100.     It  is  a 


APPENDIX  1057 

weighted  arithmetic  average  of  the  outputs  of  iron  ore  (14  per  cent),  zinc  (2  per  cent),  lead 
(4  per  cent),  copper  (13  per  cent),  natural  gas  (5  per  cent),  cement  (2  per  cent),  rails 
(3  per  cent),  steel  ingots  and  castings  (31  per  cent),  pig  iron  (12  per  cent),  coke  (5  per  cent), 
ships  (6  per  cent),  locomotives  (3  per  cent).  The  weights  were  based  approximately  on 
value  added  by  manufacture  or  on  value  of  product  (average  of  1890  and  1900). 

(2)  Germany.     Producers'  Goods.     See  above,  Chart  VII  (2). 

(8)  Great  Britain.  Producers'  Goods.  1860-1913.  See  above  Chart  V  (2).  For 
1785  this  index  covers  only  coal,  copper  ore,  copper,  tin.  Iron  and  steel  products,  machin- 
ery, etc.  are  included  from  1787.  But  in  the  period  for  which  it  has  been  charted,  coverage 
extends  to  nine  mining,  two  iron,  steel  and  machine  industries,  six  nonferrous  metal  and 
metal  goods,  three  chemical,  one  electricity,  and  one  rubber  series. 

(4)  United  States.     Consumers'  Goods.     1867-1913.     This  index  of  consumption  of 
nondurable  consumers'  goods  has  been  compiled  by  E.  M.  Hoover.    1867-1914  =  100. 
Weights  are  roughly  proportional  to  the  values  consumed  (average  of  1880,  1890,  1900,  and 
1910).     The  commodities  included  are  sugar  (16  per  cent),  coffee  (7  per  cent),  tea(l  per 
cent),  wine  (1  per  cent),  malt  liquors  (17  per  cent),  spirits  (9  per  cent),  tobacco  (18  per  cent), 
and  wheat  flour  (31  per  cent). 

(5)  Great  Britain.     Consumers'  Goods.     1860-1913.     See  above,  Chart  V,  (2).     For 
1785,  the  index  covered  cotton  yarn,  woolen  cloth,  linen,  malt,  and  paper;  for  1787,  also, 
woolen  yarn,  silk,  and  sugar.     In  the  period  for  which  it  has  been  charted,  coverage  extends 
to  six  textile,  eight  food,  drink,  and  tobacco,  two  leather,  two  paper,  one  lumber,  and  two 
chemical  series. 

(6)  Germany.     Consumers'  Goods.     See  above,  Chart  VII,  (2). 

(7)  Series  (I)  divided  by  Series  (4). 

(8)  Series  (8)  divided  by  Series  (5). 

(9)  Series  (2)  divided  by  Series  (6). 

Series  (7),  (8),  and  (9)  are  plotted  so  that  the  average  ratio  for  the  whole  period  is 
represented  by  the  "normal,"  or  100  per  cent,  line. 

CHART  XVIII.     UNITED  STATES.    ANNUAL.    LOG  SCALE 

(1)  Nondurable  Consumers'  Goods.     See  Chart  XVII  (4). 

(2)  Crop  Production.     See  Chart  VI  (5). 

(3)  Production  of  Capital  Equipment.     See  Chart  XVII  (1). 

(4)  Basic  Production  (excluding  "products  of  the  soil").     1870-1913.     This  is  Mr. 
Carl  Snyder's  index  taken  by  permission  from  G.  F.  Warren  and  F.  A.  Pearson,  The 
Physical  Volume  of  Production  in  the  United  States,   Cornell   University  Agricultural 
Experiment  Station  Memoir  144,  Ithaca,  1932,  pp.  63-64.     It  includes  coal,  pig  iron,  copper, 
zinc,  tin,  lead,  steel,  silver,  petroleum,  and  nickel. 

CHART  XIX.     GERMAN  PREWAR  PULSE  IN  RATES  OF  PERCENTAGE 
CHANGE.    ARITHMETIC  SCALE 

For  a  description  of  the  series  see  Chart  VII.  For  the  method  of  computing  the  rates 
of  percentage  change  see  Chart  IV.  The  rate  of  percentage  change  of  the  discount  rate  is 
charted  on  a  scale  one-tenth  as  large  as  that  used  for  the  other  curves. 

CHART  XX.    INTEREST  RATES  AND  UNEMPLOYMENT  PERCENTAGE. 

LOG  SCALE 

(1)  United  States.     See  Chart  VI  (2). 

(2)  Great  Britain.     See  Chart  V  (1). 

(3)  Germany.     See  Chart  VII  (3). 


1058  BUSINESS  CYCLES 

(4)  Great  Britain.  Unemployment  Percentage.  1851-1913.  The  series  was  taker 
by  permission  from  A.  C.  Pigou,  Industrial  Fluctuations,  published  by  Macmillan  &  Com- 
pany, Ltd.,  London,  1st  ed.,  1927,  pp.  353-354.  The  figures  represent  the  trade  unions 
unemployment  percentage,  as  published  in  British  and  Foreign  Trade  and  Industry 
(Second  Series),  Cmd.  2337,  pp.  89-92,  and  the  Seventeenth  Abstract  of  Labour  Statistics 
p.  2.  Persons  on  strike,  or  locked  out,  sick,  or  superannuated  are  excluded.  The  percent- 
ages for  some  of  the  earlier  years  were  partly  computed  from  expenditure  on  unemployment 
benefit. 

CHART  XXI.    UNITED  STATES  INDIVIDUAL  PRICES*  DIVIDED 
BY  THE  GENERAL  PRICE  LEVEL.    ANNUAL.    LOG  SCALE 

(1)  Wheat.     1866-1913.     Average  farm  price  per  bushel,  taken  from  the  Department 
of  Agriculture,  Yearbook  of  Agriculture,  1914,  p.  522. 

(2)  Rubber,  Para.     1856-1913.     From  1856  to  1889  the  July  items  from  the  Aldricl 
Report  [see  Chart  XI  (1)]  Part  II,  pp.  291-292.     From  1890  to  1913,  figures  of  the  Bureau 
of  Labor  Statistics,  Bulletin  149,  p.  175. 

(3)  Petroleum,  Crude,  Barreled.     1862-1913.     From  1862  to  1892  the  July  items  from 
the  Aldrich  Report,  Part  IV,  pp.  1835-1836.     These  are  monthly  averages  from  1862  tc 
1873  except  for  the  year  1867.     From  1890  to  1913  figures  of  the  Bureau  of  Labor  Statistics, 
Bulletin  390,  pp.  134-135:  annual  average  (Pennsylvania)  price  relatives  on  the  base 
1913  =  100.     The  items  up  to  1890  were  shifted  to  the  same  price  relative  basis. 

(4)  Bituminous  Coal.     1857-1913.     From  1857  to  1891  the  July  items  from  the  Aldricl 
Report,  Part  II,  p.  178.     From  1890  to  1913  annual  average  price  relatives  (1913  =  100,  oi 
"Georges  Creek,"  f.o.b.  New  York  harbor  for  1890  to  1912  and  "Pocahontas"  mine  run, 
f.o.b.  Norfolk,  Virginia,  for  1913)  from  the  Bureau  of  Labor  Statistics,  Bulletin  390,  pp, 
130-131.     Items  before  1890  were  shifted  to  the  same  price  relative  basis. 

(5)  Bessemer  Steel  Rails.     1867-1913.     From  1867  to  1889,  as  published  in  the  Aldrict 
Report,  Part  II,  p.  215.     From  1890  to  1913,  figures  of  the  Bureau  of  Labor  Statistics, 
Bulletin  149,  p.  149. 

(6)  Anthracite  Coal,  Stove.     1840-1913.     From  1840  to  1891  July  items  from  the 
Aldrich  Report,  Part  II,  p.  177.     From  1890  to  1913  figures  of  the  Bureau  of  Labor  Statis- 
tics, Bulletin  390,  pp.  126-127  ("New  York,  Tidewater"). 

(7)  Copper,  Ingot.     1840-1913.     From   1840  to  1889  July  items  from  the  Aldrict 
Report,  Part  II,  p.  185.     From  1890  to  1913  annual  average  price  relatives,  1913  =  IOC 
from  the  Bureau  of  Labor  Statistics,  Bulletin  390,  p.  150  (up  to  1907,  "lake,"  and  after, 
"electrolytic"  copper).     Earlier  items  were  shifted  to  the  same  price  relative  basis. 

(8)  Railroad  Freight  Receipts.     1852-1913.     Average  freight  receipts  per  ton-mile 
From  1852  to  1892  an  average  of  data  for  a  varying  number  of  railroads  given  in  the  Aldrict 
Report,  Part  I,  pp.  615-617.     From  1889  to  1913  average  for  years  ending  June  30,  pub- 
lished by  the  Interstate  Commerce  Commission,  Annual  Report  on  the  Statistics  of  Railwayt 
of  the  United  States.     The  two  series  were  spliced  together.     The  1908-1912  figures  are  not 
strictly  comparable  with  the  rest,  because  they  include  returns  from  switching  and  terminal 
companies. 

(9)  Pig  Iron,  No.  1  Anthracite  Foundry.     1844-1913.     From  1844  to  1890  the  data 
were  taken  by  permission  from  J.  M.  Swank,  History  of  the  Manufacture  of  Iron  in  All 
Ages,  Philadelphia,  1892,  p.  514.     From  1890  to  1913  the  figures  of  the  Bureau  of  Laboi 
Statistics,  Bulletin  390,  pp.  138-139,  have  been  used. 

Each  of  the  above  series  was  divided  by  the  wholesale  price  index,  see  Chart  IV  (1), 
Gold  or  paper  price  indices  were  used  according  to  the  nature  of  the  individual  price 
quotations. 


APPENDIX  1059 


CHART  XXII.    BRITISH  SHIPPING.    LOG  SCALE 

(1)  Total  Tonnage  in  Existence  in  the  United  Kingdom,  equal  to  steam  net  tons  plus 
one-third  of  sail  net  tons  from  the  Statistical  Abstracts  of  the  United  Kingdom.     The 
increase  in  1914  must  be  discounted  because  the  provisions  of  the  Merchant  Shipping  Act  of 
1907  became  fully  operative  on  Jan.  1  of  that  year.     From  1923  on,  the  Irish  Free  State  is 
excluded.     Data  are  plotted  as  of  ends  of  years. 

(2)  Annual  Increase  in  Total  Tonnage.    First  differences  of  series  (1)  plotted  in  the 
middle  of  each  year. 

(3)  Shipbuilding  in  the  United  Kingdom.     Net  steam  tonnage  plus  one-third  of  sail 
tonnage.     Plotted  as  of  midyear.     Warships  built  for  foreign  governments  are  included 
from  1870  to  1888,  but  from  1886  on  there  is  another  series  which  excludes  them.     The  two 
series  were  spliced.     Both  are  from  Statistical  Abstracts  for  the  United  Kingdom.     In  1897 
the  engineers'  strike  put  a  stop  to  all  construction. 

(4)  Interest  Rate.     See  Chart  V  (1). 

(5)  Price  of  "New,  Ready,  Cargo  Steamer,"  7,500  tons.     Quarterly.     This  series  was 
taken  by  permission  from  Fairplay's  Weekly  Shipping  Journal,  vol.  CXXXIV,  Jan.  10, 
1935,  p.  102. 

(6)  Freight  Rate  Index.     Annual.     This  index  was  made  up  by  splicing  together 
indices  from  four  sources  on  the  basis  of  overlapping  years. 

Board  of  Trade  figures  were  used  from  1884  to  1903. 

C.  K.  Hobson  in  his  Export  of  Capital  (London,  1914),  p.  182,  continued  the  Board  of 
Trade  Index  from  1904  to  1912. 

An  index  from  1884  to  1924  was  taken  from  F.  C.  James,  Cyclical  Fluctuations  in 
Shipping  and  Shipbuilding  Industries  (University  of  Pennsylvania  thesis,  1927),  p.  78. 
This  index  was  compiled  by  Dr.  Isserliss,  Statistician  to  the  Chamber  of  Shipping,  London, 
and  was  used  with  his  permission  and  that  of  F.  Cyril  James.  It  is  based  on  E.  A.  V. 
Angier's  figures  of  ocean  freights  published  annually  in  the  statistical  number  of  Fairplay's 
Weekly  Shipping  Journal  and  in  his  Fifty  Years  of  Freights,  1869-1919,  "Fairplay," 
London,  1920.  The  figures  represent  a  weighted  average  of  freight  rates  to  and  from 
England  for  the  whole  of  each  year,  and  do  not  give  a  complete  picture  of  the  years  during 
which  freight  rates  changed.  The  engineers'  strike  in  1897  affected  rates  and  so  did,  of 
course,  the  Spanish  American  and  South  African  wars.  In  1915  rates  were  fixed  by 
government,  and  in  1917  all  tonnage  was  requisitioned. 

From  1920  to  1934  an  index  on  the  base  1920  =  100  was  taken  from  The  Statist,  London, 
Oct.  29,  1921,  and  later  numbers.  This  index  has  also  been  compiled  by  Dr.  Isserliss.  It 
is  a  geometric  mean  of  tramp  quotations  for  eight  routes  to  and  from  the  United  Kingdom. 
The  more  important  routes  are  represented  by  more  than  one  quotation. 

CHART  XXIII.     UNITED  STATES.    LOG  SCALE 

(1)  Deposits  minus  Investments  outside  New  York  City.  1890-1914.  These  are 
actual  call  date  figures  for  all  national  banks  outside  New  York  City.  The  deposits  are 
"net"  deposits,  i.e.,  individual  deposits  (see  Chart  VI  (1))  minus  clearinghouse  exchanges, 
plus  amounts  due  to  minus  amounts  due  from  other  banks.  The  "due  to"  (or  from) 
includes  amounts  due  to  reserve  agents,  other  national  banks,  state  banks,  trust  companies, 
and  savings  banks.  These  are  not  the  "net"  deposits  of  the  Comptroller  which  are  com- 
puted for  purposes  of  obtaining  reserve  ratios.  The  figures  are  from  A.  A.  Young,  op.  cit., 
pp.  8-13  [Chart  VI  (1)].  The  investments  exclude  United  States  securities  held  to  cover 
circulation  and  United  States  securities  and  other  bonds  held  to  cover  United  States 
deposits.  They  are  also  actual  call  date  figures  for  all  national  banks  outside  New  York 
City  and  have  been  taken  from  the  same  source. 


1060  BUSINESS  CYCLES 

(2)  Clearings  outside  New  York  City.  1890-1913.  These  are  annual  averages  taken 
by  permission  from  E.  Frickey,  Bank  Clearings  outside  New  York  City,  1875-1914, 
Review  of  Economic  Statistics,  vol.  VII,  1925,  p.  260.  They  are  for  seven  selected  cities. 
See  Professor  Frickey's  article  for  further  description  of  the  series. 

(8)  Output  of  Manufacturing  and  Mining  Multiplied  by  Prices.  The  index  of  manu- 
facturing and  mining  is  described  in  Chart  VI  (8).  For  the  wholesale  price  index  see  the 
Bureau  of  Labor  Statistics  Bulletin  284,  p.  131 — all-commodity  index,  1913  =  100. 

(4)  Pay  Bolls.     This  curve  is  based  on  figures  taken  by  permission  from  Real  Wages 
in  the  United  States,  1890-1926,  by  Paul  Douglas,  published  by  Houghton  Mifflin  Com- 
pany, Boston,  1930,  pp.  440  and  463.     The  series  is  the  product  of  (a)  estimated  total 
numbers  of  persons  employed  in  manufacturing  and  transportation  ancl  (&)  average  annual 
money  earnings  of  persons  employed  in  manufacturing  and  transportation.     For  a  descrip- 
tion of  the  series  the  reader  is  referred  to  Professor  Douglas's  book.     Since  both  constituents 
are  composed  of  unavoidably  rough  estimates,  the  product  of  the  two  is  clearly  a  very 
questionable  indicator  of  the  general  course  of  events,  even  apart  from  the  fact  that  it 
would  in  any  case  not  be  strictly  comparable  with  the  other  series  used  on  the  chart. 

(5)  Series  (2)  Divided  by  Series  (4).     Keeping  in  mind  what  has  been  said  about  Series 
(4),  this  curve  is  at  best  a  highly  conjectural  approximation  to  the  measure  we  should  like 
to  have. 

CHART  XXIV.    PHILADELPHIA  CLEARINGS 

This  chart  presents  the  results  of  Dr.  Georgescu's  method  applied  to  Philadelphia  Bank 
Clearings,  1878-1914.  See  Chap.  V,  p.  215. 

CHART  XXV.     UNITED  STATES.    ANNUAL.    LOG  SCALE 

(1)  Deposits  outside  New  York  City.     The  figures  are  annual  averages  of  the  call  date 
figures  of  "individual"  deposits  described  above,  Chart  VI  (1). 

(2)  Clearings  outside  New  York  City.     See  Chart  XXIII  (2). 

(3)  Pig  iron  Consumption.     See  Chart  XII  (2). 

(4)  Equipment  Production  Index.     See  Chart  XVII  (1). 

(5)  Loans  and  Discounts  outside  New  York  City.     These  are  annual  averages  of  the 
loans  and  discounts  of  all  national  banks  outside  New  York  City  reported  for  dates  of  call 
from  A.  A.  Young,  op.  cit.,  pp.  8-13,  see  Chart  VI  (1).     Overdrafts  are  included  in  loans 
and  discounts  prior  to  1898,  but  not  later. 

CHART  XXVI.     UNITED  STATES.    ANNUAL.    LOG  SCALE 

(1)  Expenditure  on  Producers'  Goods.     Series  (1),  Chart  XVII  multiplied  by  Series  (5), 
Chart  XL 

(2)  Expenditure  on  Consumers'  Goods.     Series  (4),  Chart  XVII  multiplied  by  Series 
(1),  Chart  XI. 

(3)  Outside  Clearings.     See  Chart  XXIII  (2). 

(4)  Expenditure   on   Producers'   Goods   Divided  by   Outside   Clearings.     Series    (1) 
divided  by  Series  (3). 

(5)  Expenditure  on  Consumers'   Goods  Divided  by  Outside  Clearings.     Series   (2) 
divided  by  Series  (8). 

CHART  XXVII.     UNITED  KINGDOM.    LOG  SCALE 

(1)  Production.    See  Chart  V  (2). 

(2)  Aggregate  Money  Wage  Bill.     Annual.     From  1860  to  1901  the  series  is  taken  from 
A.  L.  Bowley,  Tests  of  National  Progress,  Economic  Journal,  vol.  XIV,  September  1904, 
by  permission  of  the  author  and  Messrs.  Macmillan  &  Company,  Ltd.,  London.     Professor 


APPENDIX  1061 

Bowley  has  since  published  a  revised  set  of  figures;  see  his  recent  book  on  Wages  and 
Income  in  the  United  Kingdom  since  1860  (Cambridge  University  Press,  1937).  From 
1901  to  1913  the  series  was  taken  by  permission  from  A.  C.  Pigou,  Industrial  Fluctuations, 
2d  ed.;  Macmillan  &  Company,  Ltd.,  London, .  1929,  pp.  383-384.  These  figures  are 
based  on  Bowley's  rates  of  wages  and  estimates  of  variations  in  the  number  of  the  wage 
earning  population. 

(3)  Provincial    Clearings.     1887-1913.     Quarterly    averages    of    monthly    data    for 
Manchester  and  Birmingham  taken  by  permission  from  D.  S.  Thomas,  An  Index  of  British 
Business  Cycles,  Journal  of  the  American  Statistical  Association,  vol.  XXI,  March  1926, 
p.  61.     The  source  quoted  is  the  Bankers'  Magazine,  London. 

(4)  Production  Multiplied  by  Prices.    Series   (1)   above  multiplied  by  Series   (3), 
Chart  V. 

(5)  Total  Taxable  Income,  adjusted  for  changes  in  the  method  of  assessment.     1842- 
1913.     Annual.     The  smoothing  effect  of  taxable  income  from  profits  being  defined  as  an 
average  of  actual  profits  [see  below,  Chart  XXVIII  (3)]  should  be  borne  in  mind.     The 
items  of  the  series  become  strictly  comparable  from  1894  on  (160  pounds  exemption  limit 
and  repairs  allowance) .    The  series  was  taken  by  permission  from  J.  Stamp,  British  Incomes 
and  Property,  pp.  31&-319,  published  by  P.  S.  King  &  Son,  Ltd.,  London,  1916. 

CHART  XXVIII.    UNITED  KINGDOM.    ANNUAL.    LOG  SCALE 

(1)  Wage  Kates.     The  series  from  1850  to  1903  has  been  taken  by  permission  from 
G.  H.  Wood,  Real  Wages  and  the  Standard  of  Comfort  since  1850,  Journal  of  the  Royal 
Statistical  Society,  vol.  LXXII,  March  1909,  pp.  99-103.     Mr.  Wood's  index  of  money 
wages  is  partly  based  on  Professor  Bowley's  work  and  partly  on  additional  material.     The 
index  is  a  weighted  average  allowing  for  changes  in  the  numbers  employed  in  the  various 
industries  (agriculture,  building,  printing,  shipbuilding,  engineering,  coal,  puddling,  cotton, 
wool  and  worsted,  gas,  and  furniture).     The  series  has  been  continued,  by  means  of  the 
figures  of  the  Board  of  Trade,  by  W.  T.  Layton  and  G.  Crowther,  An  Introduction  to  the 
Study  of  Prices,  published  by  Macmillan  &  Company,  Ltd.,  London,  1936,  from  which 
the  relevant  items  have  been  taken  by  permission. 

(2)  Wages  Bill.     See  Chart  XXVII  (2). 

(3)  Profits.     Gross  Schedule  D  assessments  (Business  Profits).     In  1908-1909  roughly 
three-quarters  of  the  assessments  were  based  on  the  average  of  the  three  preceding  years. 
As  a  rough  approximation,  therefore,  the  series  (except  in  the  cases  of  canals,  railroads, 
ironworks,  and  gasworks,  all  of  which  were  assessed  for  profits  of  the  preceding  year,  and  of 
mines,  which  were  assessed  for  the  average  of  the  five  preceding  years,  but  for  which  detailed 
annual  figures  are  given  by  Stamp  so  that  individual  adjustments  were  possible)  was  lagged 
by  two  years  and  items  were  plotted  for  midyear  instead  of  for  April.     All  figures  were 
taken  by  permission  from  J.  C.  Stamp,  op.  cit  [see  above,  Chart  XXVII  (5)  ]. 

(4)  Unemployment.     See  Chart  XX  (4). 

(5)  Wage  Bill  Divided  by  Wholesale  Prices.     For  wage  bill  see  Chart  XXVII  (2) .     The 
figures  for  wholesale  prices  were  taken  by  permission  from  G.  F.  Warren  and  F.  A.  Pearson, 
Prices,  published  by  John  Wiley  &  Sons,  Inc.,  New  York.     Their  index  joins  the  Sauerbeck 
and  Statist  indices  referred  to  in  Chart  IV  (3). 

(6)  Real  Wages.     Same  sources  as  above,  series  (1).     In  estimating  real  wages  Mr. 
Wood  assumed  that  four-fifths  of  total  wages  were  in  1850  spent  on  commodities  other  than 
housing  and  that  the  latter  item  thence  increased  steadily  (one-half  of  the  increase  in  rent 
being  attributed  to  improving  quality  of  housing  and  one-half  to  other  causes).     For  the 
rest  Mr.  Wood  took  the  unweighted  average  of  prices  of  all  commodities  of  ordinary  con- 
sumption for  which  series  were  then  obtainable.     The  index  has  been  continued  on  Board 
of  Trade  figures. 


106&  BUSINESS  CYCLES 

(7)  Real  Wages  Allowing  for  Unemployment.  Series  (6)  was  corrected  by  Mr.  Wood 
(op.  cit.,  Series  (1))  by  using  the  trade-union  unemployment  percentage  (Series  (4)).  From 
1896  on  this  series  does  not  agree  with  Pigou's  [see  Chart  XX  (4)]  or  that  given  by  the 
Abstract  of  Labour  Statistics,  which  has  been  used  instead. 

CHART  XXIX.    UNITED  STATES.    LOG  SCALE 

(1)  Wage  Rate.     1840-1913.     Annual.     Index  of  rate  per  hour,  excluding  agriculture, 
1910-1914  =  100,  during  the  Civil  War  on  currency  basis.     It  is  said  to  be  based  on  "all 
available  material."     No  allowance  has  been  made  for  the  reduction  in  hours  worked  per 
week.     The  source  is  the  Monthly  Labor  Review,  vol.  XXXII,  No.  2,  IJebruary  1931,  p.  143. 

(2)  Railroad  Earnings.     1866-1913.     Monthly  figures  corrected  for  seasonal.     Gross 
earnings  of  14  systems  as  they  existed  in  1914,  made  as  homogeneous  as  possible  by  tracing 
back  figures  for  roads  later  leased  or  absorbed  by  these  systems.     For  a  further  description 
see  A  Monthly  Index  of  Railroad  Earnings,  1866-1914  by  Arthur  H.  Cole,  Review  of 
Economic  Statistics,  vol.  XVIII,  February  1936.     Professor  Cole  was  kind  enough  to  supply 
the  figures  from  his  files. 

(3)  Dividends.     1902-1913.     Monthly.     Payments  by  industrial  corporations,  from 
the  Review  of  Economic  Statistics,  Preliminary  vol.  I,  1919,  p.  164.     The  New  York  Journal 
of  Commerce  and  Commercial  Bulletin  is  the  original  source. 

(4)  Wage  Bill.     See  Chart  XXIII  (4). 

(5)  Employment  in  Massachusetts  Factories.     1889-1913.     The  index,  1914  =  100, 
was  taken  by  permission  from  Ralph  G.  Hurlin,  Three  Decades  of  Employment  Fluctua- 
tions, the  Annalist,  vol.  XVIII  (Oct.  24,  1921),  pp.  387-388.     Obviously,  Massachusetts 
figures  cannot  be  relied  on  to  depict  national  employment  faithfully. 

(6)  Commercial   Failures.     1857-1913.     Aggregate   liabilities.     From   the   Statistical 
Abstract  of  the  United  States  for  1914,  p.  681. 

(7)  Wholesale  Prices.     See  Chart  IV  (1). 

CHART  XXX.    GERMANY.    LOG  SCALE 

(1)  Wage  Rates.     1850-1913.     Pfennigs  per  man-hour  paid  to  miners  in  the  Ruhr 
Valley,  Dortmund  Mining  District,  taken  by  permission  from  Ernst  Wagemann,  Economic 
Rhythm,  translated  from  the  German  by  D.  H.  Blelloch  and  published  by  the  McGraw-Hill 
Book  Company,  Inc.,  New  York,  p.  265. 

(2)  Wages  Bill.     Saxony.     The  figures  were  taken  by  permission  from  the  Vierteljahrs- 
hefte  zur  Konjunkturforschung,  Erganzungsheft  vol.  II  No.  3,  1927,  p.  33,  edited  by  Pro- 
fessor Wagemann.     Conditions  in  the  Saxon  Kingdom  were  considered  as  sufficiently 
typical  of  German  conditions  in  general. 

(3)  Wholesale   Prices.     1850-1913.     The   figures   were   taken   from    Wirtschaft   und 
Statistik,  vol.  V  Sonderheft  1,  1925,  p.  19,  and  the  Statistisches  Jahrbuch,  vol.  XLVII,  1926, 
p.  263.     Until  1878,  the  source  is  Adolf  Soetbeer,  Materialien  zur  Erlauterung  und  Beur- 
teilung  der  Wirtschaftlichen  Edelmetallverhaltnisse  und  der  Wahrungsfrage,  Berlin,  1886. 

(4)  Profits  in  Saxony.     From  the  article  quoted  in  Series  (2). 

(5)  Unemployment.     1906-1914.     End  of  month  figures  of  industrial  unemployment 
taken  by  permission  from  Ernst  Wagemann,  Konjunkturlehre,  published  by  R.  Robbing, 
Berlin,  1928,  p.  195. 

(6)  Series  (1)  Divided  by  Series  (3). 

CHART  XXXI.    UNITED  KINGDOM.    ARITHMETIC  SCALE 

(1)  Real  Wages,  Allowing  for  Unemployment.     See  Chart  XXVIII  (7). 

(2)  Nine-year  Moving  Average  of  (1). 

(3)  Deviations  from  nine-year  moving  average. 


APPENDIX  1063 


CHART  XXXII.    GERMANY.    ANNUAL.    LOG  SCALE 

(1)  Loans  and  Discounts.     Nine  Berlin  Grossbanken.     End  of  year  figures. 

(2)  Total  Deposits  (in  the  English  sense  of  the  term).     Nine  Berlin  Grossbanken. 
End  of  year  figures. 

(8)  Note  Circulation  of  all  German  Note-issuing  Banks. 

All  three  series  are  based  on  the  official  figures.  For  a  full  presentation  and  discussion 
of  these  and  related  data  see  L.  A.  Hahn,  Zur  Frage  des  volkswirtschaftlichen  Erkennt- 
nisgehalts  der  Bankbilanzziffern,  Vierteljahrshefte  zur  Konjunkturforschung,  vol.  I,  1926, 
Erganzungsheft  4. 

CHART  XXXIII.    UNITED  STATES.    LOG  SCALE 

(1)  Individual  Deposits  plus  Circulation  outside  New  York  City.     See  Chart  VI  (1). 

(2)  Loans  and  Discounts  outside  New  York  City.     See  Chart  XXV  (5). 

(3)  Individual  Deposits  plus  Circulation  minus  Investments  outside  New  York  City. 
Series  (1)  minus  the  investment  series  described  in  Chart  XXIII  (1). 

Actual  call  date  figures  have  been  used  in  all  three  cases. 

CHART  XXXIV.     UNITED  STATES.    LOG  SCALE 

(1)  Production  of  Industrial  Equipment.     See  Chart  XVII  (1). 

(2)  Building  Permits.     See  Chart  XIV  (5). 

(8)  Investments.     National    Banks   outside    New    York    City — a    somewhat 
sample.     See  Chart  XXIII  (1).     A  five-item  moving  total  was  plotted. 

(4)  New  Security  Listings.     Stocks  and  bonds  listed  on  the  New  York  Stock  Exchange 
minus  old  issues  and  issues  replacing  existing  securities.     The  series  was  taken  from  the 
Commercial  and  Financial  Chronicle,  by  permission  of  the  William  B.  Dana  Company, 
New  York,  and  from  the  Financial  Review. 

(5)  Loans  and  Discounts.     National  Banks  outside  New  York  City.     See  Chart  XXV 
(5).     A  five-item  moving  total  was  plotted. 

CHART  XXXV.     BANK  OF  ENGLAND  FIGURES,  ETC.     1844-1914. 
LOG  SCALE 

All  series  except  (2)  and  (3)  are  annual  averages  taken  by  permission  from  R.  H.  I. 
Palgrave,  Bank  Rate  and  the  Money  Market,  published  by  John  Murray,  London,  pp. 
12-13,  for  the  years  up  to  1900,  and  from  the  Bankers'  Almanac  and  Yearbook,  1935-1936, 
for  the  years  1900-1914. 

(1)  Private  Deposits  at  the  Bank  of  England.     There  is  a  break  in  1873,  owing  to  the 
fact  that  since  then  Chancery  Balances  have  been  put  under  the  head  of  government 
deposits.     The  average  amount  of  these  balances  was  1  million  pounds. 

(2)  London  Total  Clearings  Divided  by  Total  English  Deposits.     The  clearings  are 
monthly  average  clearings  per  working  day  during  June  or  December  multiplied  by  306, 
taken  by  permission  of  the  University  of  Chicago  Press,  Chicago,  from  The  Velocity  of 
Bank  Deposits  in  England,  by  Lionel  D.  Edie  and  Donald  Weaver,  Journal  of  Political 
Economy,  vol.  XXXVIII,  No.  4,  August  1930,  p.  897.     The  deposits  are  monthly  average 
total  deposits  in  the  joint  stock  and  private  banks  of  England  and  Wales  (except  the  Bank 
of  England)  on  June  30  or  Dec.  31,  taken  by  Edie  and  Weaver  from  the  Economist. 

(3)  London  Total  Clearings      See  Series  (2). 

(4)  Proportion  of  Reserve  (Banking  Department)  to  Deposits  plus  Bank  Post  Bills. 

(5)  Proportion  of  London  (Clearing)  Bankers'  Balances  to  the  Reserve  of  the  Banking 
Department  of  the  Bank  of  England, 


1064  BUSINESS  CYCLES 

(6)  London  Bankers'  Balances  at  the  Bank  of  England. 

(7)  "Other  Securities"  in  the  banking  department. 

(8)  Notes  Held  by  the  Public. 

CHART  XXXVI.    UNITED  STATES.    LOG  SCALE 

(1)  Bank  Clearings,  New  York  City.     1866-1914.     From  1866  to  1902  the  series  was 
taken  from  Ada  Matthews,  New  York  Bank  Clearings  and  Stock  Prices,  1866-1914,  Review 
of  Economic  Statistics,  vol.  VIII,  1926,  p.  188.     The  data  were  used  by  permission  of  the 
Commercial  and  Financial  Chronicle,  which  is  the  source  quoted  by  MissJMatthews.     From 
1908  to  1914  the  data  were  taken  from  the  Wall  Street  Journal  by  permission  of  Dow,  Jones 
&  Co.,  Inc.,  New  York.    Four-quarter  moving  totals  of  quarterly  totals  centered  on  the 
third  item  have  been  plotted. 

(2)  Value  of  Transactions,  New  York  Stock  Exchange.     1875-1914.     Annual.    Num- 
ber of  shares  traded  multiplied  by  average  price.     The  figures  for  1875  to  1909  were  taken 
from  the  United  States  National  Monetary  Commission,  Statistics  for  the  United  States, 
1867-1909,  p.  9,  where  the  Commercial  and  Financial  Chronicle  is  given  as  the  source,  while 
the  figures  for  1909  to  1914  came  directly  from  the  Commercial  and  Financial  Chronicle. 

(8)  Call  Loan  Rate,  New  York  Stock  Exchange.  1866-1914.  Moving  12-month 
totals  (centered  on  the  seventh  month)  of  monthly  averages  of  daily  renewal  rates.  The 
series  was  taken  by  permission  from  the  Standard  Trade  and  Securities  Service,  Standard 
Statistical  Bulletin,  Base  Book,  January  1982,  p.  42,  and  April  1934,  p.  6.  The  sources 
quoted  there  are  Ogle,  Dunn  &  Company,  whose  permission  to  use  their  material  is  herewith 
acknowledged,  the  Review  of  Economic  Statistics,  and  the  Financial  Review. 

(4)  Loans  and  Discounts.     New  York  City.     1870-1914.     See  Chart  XXV  (5).     Five- 
item  moving  totals  of  call  date  figures  have  been  plotted. 

(5)  Deposits.     New  York  City.     1870-1914.     "  Net "  deposits  of  national  banks.     For 
definition  of  "net"  deposits  and  also  for  the  source  see  Chart  XXIII  (1).     Five-item  mov- 
ing totals  of  call  date  figures  have  been  plotted. 

(6)  Railroad  Stock  Prices.     Monthly.     From  1854  to  1882  the  series  was  taken  by 
permission  from  the  Annalist,  vol.  XL,  Oct.  28,  1932),  p.  580.     It  is  the  Clement  Burgess 
index  of  stock  prices,  average  of  high  and  low  figures,  adjusted  for  stock  dividends.     From 
1888  to  1935  the  index  of  20  railroads — weighted  by  the  number  of  shares  of  stock  out- 
standing— of  the  Standard  Trade  and  Securities  Service,  Standard   Statistical  Bulletin, 
April  1934,  p.  30,  has  been  used  by  permission.     The  two  series  were  spliced  together. 

(7)  Industrial    Stock   Prices.     Monthly.     From    1883    to    1900   the    Axe-Houghton 
weighted  average  of  10  industrial  stocks  has  been  used  by  permission  of  the  Annalist, 
vol.  XXXVII,  Jan.  16,  1931,  p.  177.     From  1900  to  1913  the  index  used  was  taken  by 
permission  from  the  Standard  Trade  and  Securities  Service,  Standard  Statistical  Bulletin, 
April  1934,  p.  30.     It  is  composed  of  the  prices  of  50  industrial  stocks  weighted  by  value  of 
shares  outstanding,  corrected  for  rights,  stock  dividends,  changes  in  par  value,  and  con- 
solidations.    Both  indices  are  averages  of  high  and  low.     They  were  spliced  together. 

CHART  XXXVII.    GREAT  BRITAIN.    LOG  SCALE 

(1)  British  Stock  Prices.  1867-1914.  End  of  month  data;  1890  =  100;  unweighted 
arithmetic  average.  The  index  was  taken  by  permission  from  the  London  and  Cambridge 
Economic  Service,  Special  Memo.  87,  An  Index  Number  of  Securities,  1867-1914,  by  K.  C. 
Smith  and  F.  C.  Home.  The  stocks  included  divide  up  into  the  following  divisions: 
(1)  coal,  iron,  etc;  (2)  electrical  goods;  (3)  textiles;  (4)  food;  (5)  drink;  (6)  building 
materials;  (7)  lighting;  (8)  chemicals;  (9)  stores;  (10)  miscellaneous;  (11)  transport  and 
communication. 


APPENDIX  1065 

(2)  London  Call  Rate.     1888-1914.     Annual  average  rate  on  money  at  call  or  short 
notice  taken  until  1906  from  the  United  States  National  Monetary  Commission's  Statistics 
for  Great  Britain,  Germany,  and  France,  1867-1909,  p.  143.     From  1907  to  1914  the 
figures  are  annual  averages  of  rates  on  floating  money  as  published  by  the  Economist. 

(3)  London  Clearings.     1870-1914.     Total  amounts  cleared  from  1870  to  1902;  after 
that,  "town  clearings."     Both  series  are  from  the  Statistical  Abstract  for  the  United  Kingdom 
and  were  spliced  together. 

CHART  XXXVIII.    GERMANY.    LOG  SCALE 

(1)  Stock  Prices.     1870-1913.     Monthly.     This  series  was  taken,  by  permission,  from 
Otto  Donner,  Die  Kursbildung  am  Aktienmarkt,  Vierteljahrshefte  zur  Konjunkturforschung, 
Sonderheft  36,  1934,  p.  98.     It  is  composed  of  quotations  for  an  increasing  number  of  com- 
panies, 70  of  which  were  available  as  far  back  as  1890,  and  includes  banks,  shipping  com- 
panies, railways,  mines,  electric  companies,  manufacturing,  and  building. 

(2)  Industrial  Bond  Issues.     1883-1913.     Annual.     Domestic  issues  (market  value) 
from  H.  Kleiner,  Die  Emissions-Statistik  in  Deutschland,  Miinchner  Volkswirtschaftliche 
Studien  131, 1914,  Table  I,  pp.  119-124.     The  main  source  of  the  material  embodied  in  that 
table  was  Der  Deutsche  Okonomist.     The  data  compiled  by  this  periodical  refer  to  listings  at 
the  stock  exchanges  rather  than  to  issues  and  are  not  quite  complete,  especially  as  regards 
the  listings  at  the  smaller  stock  exchanges.     In  these  and  other  respects  other  compila- 
tions, such  as  those  of  the  Frankfurter  Zeitung,  the  Imperial  Statistical  office  (since  1897), 
or  the  statistical  department  of  the  Reichsbank  may  be  preferable.     But  the  material  of  the 
Deutsche  Okonomist  yields  the  longest  homogeneous  series,  and  on  the  whole  it  seemed  best 
to  use  it. 

(3)  Industrial  Stock  Issues.     From  the  same  source  as  (2). 

(4)  Foundations.     1871-1913.     Annual.     " Nominal"  capital  of  newly  founded  joint 
stock  companies,  from  Der  Deutsche  Okonomist,  vol.  XXVI,  1908,  p.  28,  and  vol.  XXXII, 
1914,  p.  412. 

(5)  Dividends,  Per  Cent  of  "Nominal"  Capital.     1870-1913.     Annual.    This  series 
has  been  taken,  by  permission,  from  the  study  by  Otto  Donner,  see  above,  Series  (1). 
Data  are  for  the  industrial  companies  included  in  the  stock  price  index  of  the  Statistisches 
Reichsamt.     From  1890  on,  the  percentages  are  weighted  according  to  the  share  capital 
of  the  various  companies. 

CHART  XXXIX.    UNITED  STATES  POSTWAR  PULSE.    LOG  SCALE 

(1)  Deposits  plus  Circulation.  The  deposits  are  net  demand  deposits  of  weekly  report- 
ing member  banks  in  100  leading  cities  outside  New  York.  The  monthly  average  figures 
used  (as  revised  in  1929)  were  supplied  to  the  author  from  the  files  of  the  Harvard  Economic 
Society  (now  Committee  on  Research  in  the  Trade  Cycle)  This  series,  subsequently 
reported  through  Aug.  23, 1935,  was  at  the  time  of  writing  available  only  through  February 
1933.  Therefore,  it  was  spliced  to  a  series  which  covered  89  cities  from  January  1932,  and 
90  cities  from  January  1934.  Since  the  Banking  Act  of  1935  changed  the  definition  of  net 
demand  deposits,  it  should  be  observed  that  that  term  is  being  used  in  its  old  sense  here: 
it  includes  all  demand  deposits  minus  United  States  government  deposits,  but  balances  due 
from  banks  and  trust  companies  subject  to  immediate  withdrawal  and  cash  items  in  process 
of  collection  were  (for  each  bank  separately)  deducted  from  balances  due  to  other  banks 
and  subject  to  immediate  withdrawal,  so  that  it  may  still  be  said  that,  very  roughly,  these 
figures  indicate  the  variations  in  the  demand  deposits  held  by  the  "public." 

The  series  of  money  in  circulation  was  taken,  by  permission,  from  J.  W.  Angell,  The 
Behavior  of  Money,  pp.  178-179,  published  by  the  McGraw-Hill  Book  Company,  Inc., 
New  York,  1936.  It  represents  a  monthly  estimate  of  currency  issued  and  not  yet  redeemed 


1066  BUSINESS  CYCLES 

minus  currency  held  in  the  Treasury  (as  asset),  in  federal  reserve  banks  or  with  federal 
reserve  agents,  and  vault  cash  of  all  reporting  banks  (as  reported  to  the  comptroller  of  the 
currency).  The  figures  for  vault  cash  are  available  only  for  June  30  of  each  year,  and  had 
to  be  estimated  from  call  date  information  for  all  member  banks'  vault  cash.  Monthly 
data  exist,  however,  for  the  sum  of  currency  in  outside  circulation  plus  vault  cash  of  all 
banks.  Since  addition  of  those  figures  for  deposits  and  circulation  would  have  given  much 
too  great  a  weight  to  circulation,  the  proportion  was  calculated  of  net  demand  deposits  of 
the  above  banks  to  total  "circulating  deposits"  as  estimated  by  Angell,  op.  tit.  It  turned 
out  to  be  fairly  steady,  the  average  for  the  year  1926  being  87.6  per  cent.  On  obviously 
very  simplifying  and  bold  assumptions,  therefore,  38  per  cent  of  that  circulation  was  added 
to  those  net  demand  deposits  of  weekly  reporting  member  banks.  It  is  impossible  to  use 
Angell's  "circulating  deposits,"  since  they  include  New  York  City  and  also  United  States 
deposits. 

(2)  Interest  Rate.     Monthly.     From  1919  to  1931,  New  York  prime  commercial  paper 
rate,  taken  by  permission  from  the  Standard  Trade  and  Securities  Service,   Standard 
Statistical  Bulletin,  Base  Book,  January  1932.     From  1932  to  1934,  the  rate  on  four  to  six 
months'  paper  has  been  taken  from  the  Survey  of  Current  Business. 

(3)  Production  Index.     Monthly.     This  is  the  Federal  Reserve  Board's  seasonally 
adjusted  index  of  manufacturing  and  mining,  taken  from  the  Federal  Reserve  Bulletin, 
vol.  XIX,  September  1933,  p.  584,  and  vol.  XXI,  May  1935,  p.  282.     It  is  a  weighted 
average  per  working  day.     The  1923-1925  average  value  was  used  for  weighting  in  the 
case  of  mining,  and  the  1923  value  added  by  manufacture  in  the  case  of  manufacturing. 
For  a  further  description  of  the  index,  see  the  Federal  Reserve  Bulletins,  vol.  XIII,  No.  2, 
February  1927;  vol.  XIII,  No.  3,  March  1927;  and  vol.  XVIII,  No.  3,  March  1932. 

(4)  Wholesale  Prices.     Monthly.     This  is  the  Bureau  of  Labor  Statistics  index  on  the 
base  1910-1914  =  100.     The  new  index  has  been  used  for  the  years  1932-1934  by  shifting 
its  base  (1926)  to  the  base  1910-1914. 

CHART  XL.    GERMAN  POSTWAR  PULSE,    LOG  SCALE 

(1)  Wholesale   Prices.     1919-1934.     Monthly   averages.     This   is   the   index   of  the 
Statistisches  Reichsamt,  taken  from  Wirtschaft  und  Statistic  (1913  =  100).     From  1919  to 
1924  this  is  an  arithmetic  average  of  prices  of  less  than  50  commodities,  essentially  raw 
materials   and   semi-manufactured   goods,    weighted   according  to  prewar  consumption. 
From  1924  to  1934  it  is  an  arithmetic  average  of  about  400  commodities  (including  finished 
products)  on  the  basis  of  quotations  increasing  from  800  to  1,000,  weighted  by  an  average  of 
prewar  and  postwar  consumption  (where  postwar  consumption  approaches  prewar  con- 
sumption) or  on  1925  consumption  figures.     Groups  are  weighted  as  follows:  agricultural 
commodities  35  per  cent,  imported  groceries  3  per  cent,  industrial  raw  materials  and  semi- 
manufactured products  88  per  cent,  manufactured  goods  24  per  cent.     The  two  indices 
were  spliced  together.     For  more  detailed  information  see  Wirtschaft  und  Statistik,  vol.  VI, 
1926,  p.  875,  and  Vierteljahrshefte  zur  Statistik  des  Deutschen  Reiches,  vol.  XXXVI,  1927, 
p.  37,  and  vol.  XLI,  1932,  p.  139. 

(2)  Interest  Rate.     Rate  on  bank  acceptances,  monthly  average.     Taken  by  permis- 
sion from  the  Institut  fiir  Konjunkturforschung,  Konjunktur-Statistisches  Handbuch,  1936, 
p.  113. 

(3)  Production.     Quarterly  index  of  industrial  production  per  working  day  (1928  == 
100),  taken  by  permission  from  the  Institut  fttr  Konjunkturforschung,  Konjunktur-Statis- 
tisches Handbuch,  1936,  p.  52.     Seasonal  variations  are  eliminated.     The  Saar  is  included 
from  March  1935,  so  that  the  series  is  not  strictly  homogeneous.     The  index  was  first 
calculated  in  1927,  and  revised  in  1929,  1931,  1933,  and  1935.     It  is  an  arithmetic  average. 
Series  are  weighted  within  each  group  by  value  added,  number  employed,  and  horse-power 


APPENDIX  1067 

installed.  Groups  are  separately  weighted  by  value  added.  The  index  now  represents 
66  per  cent  of  industrial  net  production.  For  further  discussion,  see  Vierteljahrshefte  zur 
Konjunkturforschung,  vol.  IV,  No.  4A,  1930;  vol.  VI,  No.  1A,  1931;  vol.  VII,  No.  4A,  1933; 
and  the  Wochenbericht  of  the  Institut  filr  Konjunkturforschung,  vol.  VIII,  No.  24,  for 
June  19,  1935. 

(4)  Deposits  and  Circulation.  The  deposits  (Glaubiger,  mostly  what  in  this  country 
would  be  called  business  deposits)  are  those  of  five  big  banks  (evidently  taken  as  a  sample, 
but  an  unsatisfactory  one,  of  total  deposits),  end  of  month  figures  from  the  Institut  fUr 
Konjunkturforschung,  Konjunktur-Statistisches  Handbuch,  1936,  p.  136.  The  five  banks 
are:  Deutsche  Bank  und  Disconto-Gesellschaft,  Berlin;  Dresdqer  Bank,  Berlin;  Commerz- 
undPrivat  Bank,  Berlin;  Bayerische  Hypotheken-und  Wechsel-Bank,  Mlinchen;  Allgemeine 
Deutsche  Credit-Anstalt,  Leipzig.  Figures  include  the  effects  of  amalgamations.  No 
figures  are  available  for  January  of  each  year,  and  from  1925  to  1927  figures  are  given  only 
for  February,  April,  June,  August,  October,  and  December.  The  circulation  series  is  total 
money  in  circulation,  end  of  month  figures  from  the  Institut  fiir  Konjunkturforschung, 
Konjunktur-Statistisches  Handbuch,  1936,  p.  130.  No  allowance  is  made  for  money  in 
banks.  * 

CHART  XLI.     BRITISH  POSTWAR  PULSE.    LOG  SCALE 

(1)  Wholesale   Prices.     1919-1934.     Monthly.     The   index,    1913  =  100,    was   taken 
from  the  Board  of  Trade  Journal.     For  a  description  see  A.  W.  Flux,  The  Measurement  of 
Price  Changes,  Journal  of  the  Royal  Statistical  Society,  vol.  LXXXIV,  Part  2,  March  1921. 
It  is  the  geometric  average  of  the  wholesale  prices  of  150  commodities  divided  into  eight 
groups  of  approximately  equal  total  values  of  output  or,  in  the  case  of  imported  consumers' 
goods,  of  imports  in  1907.     The  basis  of  valuation  is  the  census  of  production  of  1907. 
Within  the  eight  groups,  classes  of  commodities  are  represented  by  a  number  of  price  series 
varying  according  to  the  same  criterion.     The  newly  revised  index  covering  200  commodi- 
ties and  weighting  according  to  the  values  of  the  1930  census  of  production  presents  almost 
exactly  the  same  picture  from  1930  to  1934.     Therefore,  it  was  not  thought  necessary  to 
redraw  the  curve  when  the  new  index  became  available. 

(2)  Interest  Rate.     Monthly.     Average  rate  on  three  months'  commercial  paper  for 
week  ending  the  fifteenth  of  the  month.     Data  were  taken  by  permission  from  the  monthly 
bulletins  of  the  London  and  Cambridge  Economic  Service. 

(3)  Deposits  plus  Circulation.     The  deposits  are  current  accounts  of  the  10  London 
Clearing  Banks,  average  for  the  month,  taken  from  the  Report  of  the  Committee  on  Finance 
and  Industry,  1931,  Cmd.  3897,  pp.  284-289,  and  for  later  years  from  the  Bank  of  England 
Statistical  Summary.     In  the  case  of  the  National  Bank,  Ltd.,  only  figures  relating  to 
offices  in  England  are  included.     Some  items  in  the  earlier  years  are  estimates.     The 
circulation  series  represents  Bank  of  England  and  currency  notes  in  circulation  from  the 
eleventh  to  the  seventeenth  of  each  month,  taken,  by  permission,  from  the  monthly 
bulletins  of  the  London  and  Cambridge  Economic  Service. 

(4)  Production  Index.     Quarterly  index  of  production  taken,  by  permission,  from  the 
London  and   Cambridge  Economic  Service,  Monthly  Bulletin,  vol.  VIII,  No.  4  (Special 
Quarterly  Issue,  Apr.  23,  1929)  and  later  Special  Quarterly  Issues.     This  is  an  arithmetic 
average,  1924  =  100,  of  individual  output  series,  weighted  according  to  net  output  as  given 
by  the  1924  census  of  production.     For  further  discussion  see  the  Special  Memorandum  8 
of  the  London  and  Cambridge  Economic  Service,  The  Physical  Volume  of  Production,  by 
J.  W.  F.  Rowe.     The  subindexes  are  (1)  coal  mining;  (2)  pig  iron,  steel,  shipbuilding, 
railroad  vehicles;  (3)  copper,  lead,  tin,  zinc;  (4)  cotton,  silk;  (5)  wheat,  flour,  cocoa,  tobacco; 
(6)  oil  seed  crushings,  heavy  chemicals;  (7)  paper. 


1068  BUSINESS  CYCLES 


CHART  XLII.    INDUSTRIAL  PRODUCTION.    ANNUAL.    LOG  SCALE 

(1)  United  States.     Index  taken,  by  permission,  from  the  Standard  Trade  and  Securi- 
ties Service,  Standard  Statistical  Bulletin,  December  1935,  p.  39.     A  weighted  composite  of 
64  series.     Corrected  for  seasonal.     Revised  in  1933.     For  series  and  weights  see  the 
Standard  Trade  and  Securities  Service,  Basic  Statistics,  vol.  LXXX,  No.  29,  June  5,  1936, 
p.  D-36. 

(2)  Germany.     See  Chart  VII  (2). 

(3)  Great  Britain.     Hoffmann  Index.     See  Chart  V  (2). 

(4)  Great  Britain.    London  and  Cambridge  Economic  Service  Index.     See  Chart  XLI 
(4). 

CHART  XLIII.     UNITED  STATES  PRODUCTION  SERIES.     1919-1934. 
MONTHLY.    LOG  SCALE 

All  the  series  taken  by  permission  from  Y.  S.  Leong,  Indexes  of  the  Physical  Volume  of 
Production  of  Producers'  Goods,  Consumers'  Goods,  Durable  Goods,  and  Transient  Goods, 
Journal  of  the  American  Statistical  Association,  vol.  XXX,  No.  189,  June  1935.  All  series 
are  on  a  daily  average  output  basis  and  adjusted  for  seasonal  variation.  The  1923-1925 
average  =  100.  The  aggregative  method  was  used  to  combine  the  individual  series 
(groupwise)  into  composite  index  numbers.  Weighting  is  by  value  added  by  manufacture. 
The  average  value  added  of  the  census  years  1923,  1925,  and  1927  has  been  used  for 
producers'  goods,  all  durable  goods,  and  total  manufacturing  output  from  1922,  and  for 
consumers'  goods  excluding  motorcars,  and  for  transient  goods  from  1923  on.  Those 
weights  seemed,  however,  inappropriate  for  the  preceding  years.  Accordingly,  Mr.  Leong 
computed  another  set  of  index  numbers,  using  the  value  added  figures  of  1919,  and  com- 
bined it  with  the  first  set.  The  resulting  index  is,  for  1919  to  1922,  a  simple  geometric 
average  of  the  two  in  the  case  of  consumers'  goods  excluding  motorcars,  and  of  transient 
goods,  and  for  1919  to  1921  a  geometric  average  with  variable  weights  in  the  case  of 
producers'  goods,  all  durable  goods,  and  total  manufacturing  output.  This  procedure, 
though  not  easy  to  defend  on  general  principles,  probably  represents  in  this  case  a  fair 
approximation  to  the  theory  outlined  in  Chap.  IX. 

(1)  Producers'   Goods.     Unfinished  goods  or  goods   used  to  produce  other  goods. 
Certain  textiles,  forest  products,  paper  and  printing,  chemicals,  leather,  stone  and  clay,  iron 
and  steel,  nonferrous  metals  and  their  products,  transportation  equipment. 

(2)  Consumers'  Goods.     Series  representing  fabricated  goods  for  immediate  or  nearly 
immediate  consumption.     Food  and  kindred  products,  certain  kinds  of  textiles  and  print- 
ing, gasoline  and  kerosene,  rubber  products,  shoes,  gloves,  radiators,  sanitary  ware,  auto- 
mobiles, tobacco  manufactures. 

(3)  Consumers'    Goods    Excluding    Automobiles.     Same    as    Series    (2),    except   for 
automobiles. 

(4)  Durable  Goods.     Goods  with  an  average  useful  life  of  more  than  two  years.     Forest 
products,  coke,  stone,  clay,  glass,  iron,  steel,  nonferrous  metals  and  their  products,  and 
transportation  equipment. 

(5)  Transient  Goods.     Products  with  an  average  useful  life  not  exceeding  two  years. 
Food  and  kindred  products,  textiles  and  their  products,  paper  and  printing,  chemical, 
rubber,  and  leather  products. 

(6)  Manufactures.     All  the  series  used  in  the  above  indices:  15  food  and  kindred  prod- 
ucts; 4  textiles  and  textile  products;  2  forest  products;  10  paper  and  printing;  7  chemical 
and  allied  products;  2  rubber  products;  5  leather  and  leather  products;  2  stone,  clay,  and 
glass  products;  4  iron  and  steel  products;  4  nonferrous  metals  and  their  products;  5  trans- 
portation equipment;  3  tobacco  manufactures.    These  items  represent  directly  about 


APPENDIX  1069 

50  per  cent  of  the  value  added  by  manufacture  in  all  manufacturing  industries  in  the  census 
years  1923,  1925,  and  1927. 

CHART  XLIV.     UNITED  STATES  CUSTOMERS'  LOAN  RATES.    ANNUAL. 
ARITHMETIC  SCALE 

All  three  series  are  yearly  averages  of  monthly  averages  as  reported  by  banks  to  the 
Federal  Reserve  Board.  The  monthly  averages  are  based  on  rates  reported  for  three  types 
of  customers'  loans,  commercial  loans,  and  time  and  demand  loans  on  securities.  The 
averages  are  weighted  according  to  the  relative  importance  of  each  of  these  three  types  of 
loans,  the  relative  importance  of  each  reporting  bank,  and  the  relative  importance  of  each 
city  in  each  group. 

CHART  XLV.    UNITED  STATES  PROFIT  RATIOS.    ANNUAL. 
ARITHMETIC  SCALE 

All  series  were  taken,  by  permission,  from  W.  L.  Crum,  Corporate  Earning  Power  in  the 
Current  Depression,  Business  Research  Studies  10,  Harvard  University  Graduate  School  of 
Business  Administration.  The  profit  ratio  is  statutory  net  income  (minus  federal  taxes), 
divided  by  gross  income.  The  data  are  from  federal  income  tax  reports.  For  further 
explanation,  see  the  article  mentioned. 

CHART  XLVL    UNITED  STATES.    191&-1934.    MONTHLY.    LOG  SCALE 

(1)  Outside  Debits.     Bank  debits  to  individual  deposit  accounts  in  140  cities  outside 
New  York  City  as  published  by  the  Federal  Reserve  Board.     For  the  most  part,  these 
debits  arise  from  checks  against  depositors'  accounts  and  represent  payments.     All  debits 
to  deposit  accounts  of  individuals,  firms,  corporations,  and  U.  S.  government,  county,  and 
municipal  accounts  enter  the  total,  however,  including  debits  to  war  loan  deposit  accounts, 
savings  accounts,  payments  from  trust  accounts,  and  certificates  of  deposit  paid,  except 
debits  in  settlement  of  clearinghouse  balances,  debits  to  accounts  of  other  banks,  payments 
of  cashiers'  checks,  charges  to  expenses  and  miscellaneous  accounts,  corrections,  and  the 
like.     See  the  Twenty-Second  Annual  Report  of  the  Federal  Reserve  Board,  1935,  p.  175. 

(2)  Outside   Net   Demand   plus   Time   Deposits   minus   Outside   Investments.     See 
Chart  L  (4)  and  (7). 

(3)  Index  of  Pay  Rolls  in  Manufacturing  Industry.     This  is  the  revised  index  of  the 
Bureau  of  Labor  Statistics,  see  the  Bureau's  Bulletin  610,  p.  22.     It  represents  average 
weekly  factory  pay  rolls  (1923-1925  =  100).     In  1925  54  industries  were  covered,  employ- 
ing 83  per  cent  of  the  workers  in  all  manufacturing  industries.     The  establishments  surveyed 
employed  about  50  per  cent  of  the  83  per  cent.     Since  1931  coverage  extends  to  90  indus- 
tries.    Weighting  is  according  to  pay  rolls  in  1923-1925.     The  index  is  not  corrected  for 
seasonal  variations.     Another  correction  has,  however,  been  made  which  should  be  men- 
tioned.    Since  the  index  revealed  a  downward  bias  when  compared  with  the  trend  shown  by 
the  figures  of  the  biennial  census,  it  was  adjusted  so  as  to  conform  to  that  trend.     The 
operation  consisted  essentially  in  straight-line  adjustments  to  the  averages  of  pay  rolls  for 
pairs  of  census  years:  an  appropriate  cumulative  unit  was  applied  to  the  24-month  interval, 
in  order  to  bring  up  the  average  for  each  year  to  the  amount  indicated  by  the  census  figure. 
For  example,  the  original  data  for  the  years  1919  to  1921  were  thus  adjusted  to  the  annual 
census  averages  for  those  two  years.     Then,  before  going  on  to  the  adjustment  to  the  census 
figures  from  1921  to  1923,  preliminary  adjustment  for  1921  to  1923  became  necessary  in 
order  to  reestablish  the  comparability  destroyed  by  the  previous  adjustment. 

(4)  Outside  Net  Demand  Deposits.     See  Chart  XXXIX  (1),  first  component. 

(5)  Annual  Aggregate  Realized  Income.     Morris  A.  Copeland,  in  How  Large  is  Our 
National  Income?,  Journal  of  Political  Economy,  vol  XL,  December  1932,  p.  773,  presents 


1070  BUSINESS  CYCLES 

a  series  labeled  "realized  income  revised,"  which  is  a  revision  of  Wilford  King's  series.  It 
excludes  both  undistributed  profits  and  capital  gains.  W.  L.  Crum,  in  turn,  revised  Pro- 
fessor Copeland's  series,  see  The  National  Income  and  its  Distribution,  Journal  of  the 
American  Statistical  Association*  vol.  XXXI,  March  1935,  p.  36,  deducting  "imputed 
income  yielded  by  individually  owned  urban  homes  and  other  durable  consumers'  goods" 
by  means  of  the  figures  given  in  America's  Capacity  to  Consume,  by  Maurice  Leven,  H.  G. 
Moulton,  and  Clark  Warburton,  Brookings  Institution,  1934,  p.  153. 

For  the  chart  Professor  Crum's  revised  series,  taken  by  permission  from  the  Journal  of 
the  American  Statistical  Association  was  spliced  on  to  the  series  prepared  by  the  Depart- 
ment of  Commerce  as  published  in  the  Survey  of  Current  Business,  July,  1936,  by  Robert  R. 
Nathan,  Chief,  Income  Section,  Division  of  Economic  Research.  This  is  a  continuation 
of  the  work  presented  in  National  Income  1929-1932,  Senate  Document  124,  Seventy-third 
Congress,  Second  Session.  The  figures  used  are  those  of  "national  income  paid  out." 
They  represent  payments  to,  or  receipts  by,  individuals,  such  as  wages,  salaries,  interest, 
dividends,  entrepreneurial  withdrawals,  net  rents,  and  royalties,  and  differ  from  "income 
produced"  by  positive  or  negative  business  savings. 

(6)  Outside  Debits  Divided  by  Outside  Net  Demand  plus  Time  Deposits.     See  Series 
(1)  and  Chart  L  (4). 

(7)  Hourly  Earnings  in  Manufacturing.     This  series  is  labeled  wage  rates  on  the  chart 
but,  of  course,  may  vary  in  responses  to  changes  in  the  amount  of  overtime  and  in  the 
composition  of  the  working  force.     The  figures  are  quarterly  for  24  industries,  from  1920  to 
1926,  and  monthly  for  25  industries,  from  1927  on.     Data  through  1926  were  taken  by 
permission  from  the  National  Industrial  Conference  Board,  Wages  in  the  United  States, 
1914-1930,  p.  44,  and  from  1927  to  date  from  the  Survey  of  Current  Business,  vol.  XII, 
No.  12,  December  1932,  p.  18,  and  later  numbers.     At  the  end  of  1931  over  1,400  plants 
were  covered.     The  weekly  earnings  from  which  the  hourly  rates  are  derived  are  computed 
by  weighting  the  average  weekly  earnings  in  each  industry  by  the  relative  importance  of  the 
industry  as  revealed  by  the  Census  of  Manufactures  of  1923.     Moreover,  the  weights  are 
made   to    reflect    the    relative    importance    of   each    labor   group    in   each    industry  as 
ascertained  by  the  Conference  Board's  investigations  during  1927—1929.     Data  on  hours  of 
work  are  from  the  same  number  of  plants  and  workers  as  the  weekly  earnings.     The  base 
is  1923  =  100. 

This  chart  was  made  up  before  the  National  Industrial  Conference  Board's  volume  on 
Wages,  Hours,  and  Employment  in  the  United  States,  1914-1936,  became  available.  That 
volume  carries  the  series  back  all  the  way  on  a  monthly  basis  and,  while  the  general  method 
is  the  same,  certain  revisions  in  the  details  of  procedure  have  been  made.  Results  differ, 
however,  so  insignificantly  from  the  series  above  described  that  it  was  not  thought  necessary 
to  redraw  the  chart  and  to  recalculate  the  series  into  which  these  figures  enter  [Chart  XL VII 
(6)  and  (7)]. 

CHART  XLVII.     UNITED  STATES.    LOG  SCALE 

(1)  Hourly  Wage  Rates.     See  Chart  XLVI  (7). 

(2)  Pay  Rolls.     See  Chart  XLVI  (3). 

(3)  Employment.     Manufacturing   Industries.     For   source   and  description   of  the 
material  used  see  Chart  XLVI  (3)  [Pay  rolls].     In  this  index  the  weights  are  based  on  the 
annual  average  number  of  wage  earners  employed  in  the  industry  or  group  from  1923  to 
1925. 

(4)  Cost  of  Living.     This  index  has  been  constructed  by  the  National  Industrial  Con- 
ference Board,  and  is  here  used  by  permission.     It  aims  specifically  at  the  prices  relevant 
to  the  American  wage  earner's  budget  (1923  =  100).     The  five  major  groups  (postwar 
budget  weights)  are  food  (33  per  cent),  housing  (20  per  cent),  clothing  (12  per  cent),  fuel 


APPENDIX  1071 

and  lighting  (5  per  cent),  and  sundries  (30  per  cent).  The  food  index  used  is  the  Bureau 
of  Labor  Statistics  retail  food  price  index  for  the  fifteenth  of  each  month.  From  1920  to 
1925  the  figures  are  based  on  three  comprehensive  surveys  for  each  March,  July,  and 
November,  and  less  comprehensive  information  for  the  other  months.  Since  1925,  monthly 
calculations  have  been  made  on  a  comprehensive  basis.  The  figures  were  taken  from  the 
Standard  Trade  and  Securities  Service,  Standard  Statistical  Bulletins,  and  the  U.  S.  Depart- 
ment of  Commerce,  Survey  of  Current  Business,  vol.  XVI,  No.  1,  January  1936,  p.  19.  An 
improved  index,  which  differs  from  the  one  used  sometimes  by  as  much  as  2  per  cent,  has 
been  published  since  in  the  National  Industrial  Conference  Board's  volume  on  Cost  of 
Living  in  the  United  States,  1914-1936.  It  gives  somewhat  lower  figures  for  the  earlier 
and  somewhat  higher  figures  for  the  later  years. 

(5)  Wholesale  Prices.     See  Chart  XXXIX  (4). 

(6)  Series  (1)  Divided  by  Series  (5). 

(7)  Series  (1)  Divided  by  Series  (4). 

(8)  Series  (2)  Divided  by  Series  (4). 

(9)  Series  (2)  Divided  by  Series  (5). 

CHART  XLVIII.    GERMANY.    LOG  SCALE 

(1)  Unemployment.     The  number  of  unemployed  registered  at  the  employment  offices. 
End    of    month    figures.     Data    taken    from    the    Institut    ftlr    Konjunkturforschung, 
Konjunktur-Statistisches  Handbuch,  1936.     From  March  1935  the  Saar  country  is  included. 

(2)  Wage  Rates.     Average  standard  hourly  (or  piece)  rates  of  workers  of  the  highest 
standard  age  group.     First  of  the  month  figures,  taken  from  the  Institut  filr  Konjunktur- 
forschung, Konjunktur-Statistisches  Handbuch,  1936.     The  old  index  of  the  statistical  office 
of  the  Reich  from  1925  to  1927  was  chained  to  the  new  index,  which  goes  back  to  1928. 
The  old  index  included  a  smaller  number  of  industrial  groups  and  only  male  workers. 
The  new  index  includes  female  workers  and  is  weighted  according  to  the  average  number 
employed  in  the  years  1928  to  1930.     For  further  information  sec  the  Vierteljahrshefte  zur 
Statistik  des  Deutschen  Reichs,  vol.  XL,  1931,  Heft  2,  pp.  94  et  seq. 

From  1932  to  1934  a  third  index  has  been  used.  It  represents  standard  rates  in  the 
highest  age  group  of  unskilled  male  workers,  and  was  taken  from  the  Statistisches  Jahrbuch 
des  Deutschen  Reichs,  vol.  LV,  1934,  p.  279.  It  was  spliced  on  to  the  other  index. 

Of  course,  standard  wage  rates  cannot  be  trusted  to  represent  the  price  of  labor  accu- 
rately. 

(3)  Wage  Bill.     Income  from  wages  and  salaries  according  to  the  estimate  of  the  Insti- 
tut fur  Konjunkturforschung — excluding  pensions  and  including  the  income  from  all 
emergency  and  relief  employment.     The  items  for  the  second  and  third  quarter  of  1935 
do  not  include  the  Saar.     The  figures  were  taken  from  the  Institut  flir  Konjunkturforschung, 
Konjunktur-Statistisches  Handbuch,  1936. 

(4)  Dividends.     The   data  of  the  Institut  fiir  Konjunkturforschung's  Konjunktur- 
Statistisches  Handbuch,  1936,  are  believed  to  give  a  rough  picture  of  the  course  of  total 
dividends.     Two  figures  are  presented  for  each  quarter.     One  is  comparable,  as  regards 
companies  included,  with  the  figure  given  for  that  quarter  of  the  preceding  year,  and  the 
other  is  comparable  with  the  figure  for  the  corresponding  quarter  of  the  following  year. 
Link  relatives  could  not  be  constructed  between  the  different  quarters  in  the  same  year. 
Therefore,  the  fourth  quarter  of  1926  was  used  as  the  basis  of  link  relatives  of  the  items  for 
the  fourth  quarter  of  each  year,  because  more  than  half  of  all  dividends  are  paid  out  in  the 
fourth  quarter.     However,  any  significant  change  in  the  dates  of  dividend  payments 
would  destroy  any  value  the  above-described  relatives  may  have  as  an  index  of  total 
dividends  paid  out. 


1072  BUSINESS  CYCLES 

(5)  Cost  of  Living.     Both  the  old  index  of  the  Statistisches  Reichsamt  covering  the 
years  from  1925  to  1928  and  the  new  index  constructed  from  1928  on  were  taken  from  the 
Institut  ftir  Konjunkturforschung,  Konjunktur-Statistisches  Handbuch,  1936.     From  1925 
to  1928  groups  are  weighted  mostly  on  the  basis  of  the  budget  inquiry  of  1907,  but  some 
account  was  taken  of  the  changes  which  occurred  since  the  war.     From  1928  on,  the  index  is 
weighted  according  to  the  results  of  the  budget  inquiry  of  1927-1928.     The  old  and  new 
indices  were  linked  together  by  the  chain  method. 

(6)  Wholesale  Prices.     See  Chart  XL  (1). 

(7)  Series  (2)  Divided  by  Series  (5). 

(8)  Series  (2)  Divided  by  Series  (6). 

(9)  Series  (3)  Divided  by  Series  (6). 

(10)  Series  (3)  Divided  by  Series  (5). 

CHART  XLIX.    GREAT  BRITAIN.    LOG  SCALE 

(1)  Profits.     1920-1934.     Annual.     General  index  of  the  returns  to  "industrial  capi- 
tal" as  a  whole.     Base  1924  =  100.     No  allowance  was  made  for  the  increase  in  capital 
invested  since  1924.     The  "capital"  includes  debentures  and  other  items,  the  returns  on 
which  vary  slowly.     The  data  for  1932,  1933,  and  1934  are  provisional.     The  figures  were 
taken  from  Sir  J.  C.  Stamp's  letter  to  the  London  Times,  Aug.  9,  1935,  p.  6.     Obviously 
this  series  indicates  but  very  imperfectly  the  variations  we  would  like  to  measure. 

(2)  Wage  Rates.     The  figures  were  taken  by  permission  from  A.  L.  Bowley,  A  New 
Index  Number  of  Wages,  London  and  Cambridge  Economic  Service,  Special  Memo  28. 
From  1919  to  1925  it  is  an  unweighted  index  of  wage  rates  in  11  occupations  (December 
1924  =  100).     This  was  spliced  on  to  the  new  index  of  average  weekly  wages,  which  is 
available  back  to  1925.     The  new  index  covers  20  occupations  and  is  weighted  according 
to  the  wage  bill  of  each  group  in  the  base  period.     Figures  are  for  the  fifteenth  of  the  month. 
Professor  Bowley  has  since  revised  his  figures.     For  the  new  index,  see  again  his  recent 
book,  Wages  and  Income  in  the  United  Kingdom  since  1860. 

(3)  Employment.     This  index  was  taken  from  the  supplements  to  the  Economist, 
Oct.  21,  1933,  p.  6,  and  later  numbers.     It  is  based  on  the  number  of  insured  workers  in 
employment  in  Great  Britain,  as  published  in  The  Ministry  of  Labour  Gazette  for  the  third 
week  of  each  month.     Seasonal  fluctuations  were  not  eliminated  (1924  =  100). 

(4)  Commercial  Failures.     The  annual  number  of  commercial  failures  in  England  and 
Wales,  taken  from  the  Statistical  Abstracts  for  the  United  Kingdom. 

(5)  Wholesale  Prices.     See  Chart  XLI  (1). 

(6)  Cost  of  Living.     This  index  is  an  arithmetic  average  (July  1914  =  100).     Retail 
price  quotations  (first  of  the  month)  are  obtained  from  large  and  small  towns,  and  for  a 
number  of  articles  in  each  of  the  five  major  groups.     Each  group  is  weighted  according 
to  its  importance  in  the  budget  of  a  "typical"  workingman's  household  as  ascertained  by 
the  Board  of  Trade  Study  of  2,000  workingmens'  families  made  in  1904.     The  index  has 
been  taken  from  the  Abstract  of  Labour  Statistics.     See  The  Cost  of  Living  Index  Number, 
Method  of  Compilation,  H.M.S.O. 

(7)  Series  (2)  Divided  by  Series  (5). 

(8)  Series  (2)  Divided  by  Series  (6). 

CHART  L.     UNITED  STATES.     1919-1934.     MONTHLY.    LOG  SCALE 

(1)  Net  Demand  Deposits,  New  York  City.     Monthly  averages  of  weekly  data  of 
weekly  reporting  member  banks  in  New  York  City.     Figures  as  revised  in  1929.     Federal 
Reserve  Board  data.     See  Series  (2). 

(2)  Total  Deposits — Time  and  Demand:  New  York  and  Outside  Net  Demand  plus 
Time  Deposits  of  Weekly  Reporting  Member  Banks  in  Leading  Cities.     Monthly  aver- 


APPENDIX  1073 

ages  of  weekly  figures.  Data  as  revised  in  1929  (see  Federal  Reserve  Bulletin,  vol.  XV, 
No.  1,  January  1929).  For  further  description  of  net  demands  deposits  see  Chart  XXXIX 
(1),  first  component.  Remarks  on  the  changing  number  of  cities  and  splicing  operations 
also  apply  to  Time  Deposits. 

(3)  Bank  Debits  outside  New  York  City.     See  Chart  XLVI  (3). 

(4)  Outside  Time  plus  Net  Demand  Deposits.     See  Series  (2)  above. 

(5)  Loans  and  Discounts  of  Reporting  Member  Banks  outside  New  York  City.     For 
changing  number  of  cities  and  splicing  see  Chart  XXXIX  (1),  first  component.     Monthly 
averages  as  revised  1929.     For  further  description  of  the  series,  see  the  Federal  Reserve 
Bulletin,  vol.  XV,  No.  1,  January  1929. 

(6)  Net  Demand  Deposits  outside  New  York  City.     See  Chart  XXXIX  (1). 

(7)  Outside  Investments.     Monthly  averages  of  figures  for  weekly  reporting  banks 
outside  New  York  City.     Federal  Reserve  Board  data  as  revised  in  1929.     For  number  of 
cities  and  splicing  operations  see  Chart  XXXIX  (1),  first  component.     For  further  descrip- 
tion see  the  Federal  Reserve  Bulletin,  vol.  XV,  No.  1,  January  1929. 

CHART  LI.    LONDON  CLEARING  BANKS.    LOG  SCALE 

(1)  Commercial   Bills   Discounted.     Ten  London   Clearing  Banks.     For  source   see 
Chart  XLI  (3). 

(2)  Treasury  Bills  Discounted.     See  Series  (1). 

(3)  Treasury  plus  Commercial  Bills  Discounted.     1930-1936.     These  are  figures  for 
nine  London  Clearing  Banks  (excluding  the  National  Bank,  Ltd.).     Monthly  averages, 
taken  by  permission  from  the  monthly  bulletins  of  the  London  and  Cambridge  Economic 
Service. 

(4)  Advances  (Loans  and  Overdrafts).     From  January  1919  to  March  1931,  see  Series 
(1).     The  figures  for  nine  Clearing  Banks  (see  Series  (3))  were  spliced  on  for  1931-1936. 

(5)  Investments.     From  1919  to  1931  see  Series  (1).     Monthly  averages  for  the  nine 
Clearing  Banks,  see  Series  (3),  were  spliced  on  from  1931  to  1936.     Investments  are  exclu- 
sive of  investments  in  affiliated  banks. 

(6)  Deposit  (Time)  Accounts.     From  1919  to  1931,  see  Series  (1).     From  March  1931 
on,  the  figures  were  taken  from  the  Bank  of  England  Statistical  Summary.     There  were 
minor  changes  in  1931  in  the  allocation  of  certain  deposits.     The  1936  figures  are  for  11 
banks. 

(7)  Current  (Demand)  Accounts.     Same  sources  as  Series  (6). 

(8)  Country  plus  Provincial  Clearings.     These  are  country  clearings  at  the  London 
clearinghouse  plus  clearings  at  11  provincial  clearinghouses.     From  1920  to  1928  these 
figures  were  taken  by  permission  of  the  University  of  Chicago  Press,  Chicago,  from  L.  D. 
Edie  and  D.  Weaver,  Velocity  of  Bank  Deposits  in  England,  Journal  of  Political  Economy, 
vol.  XXXVIII,  No.  4,  August  1930.     For  the  rest  of  the  period  this  series  has  been  spliced 
on  to  the  one  published  by  the  London  and  Cambridge  Economic  Service. 

(9)  Country  plus  Provincial  Clearings  Divided  by  Current  Accounts  of  Ten  London 
Clearing  Banks.     1920-1928.     The  clearings  are  those  described  in  Series  (8)  and  the 
current  accounts  are  as  estimated  by  L.  Edie  and  D.  Weaver,  op.  cit.,  p.  395.     The  reader 
is  referred  to  that  article  for  further  discussion  of  the  series  and  the  method  of  computation. 

CHART  LII.    UNITED  STATES.    LOG  SCALE 

(1)  Call  Loan  Rate,  New  York  Stock  Exchange.  Monthly  averages  of  daily  renewal 
rates.  The  series  was  taken,  by  permission,  from  the  Standard  Trade  and  Securities 
Service,  Standard  Statistical  Bulletin,  Base  Book,  January  1932  and  later  numbers.  The 
sources  quoted  there  are  Messrs.  Ogle,  Dunn  &  Company,  whose  permission  is  herewith 
acknowledged,  and  the  Federal  Reserve  Board.  See  Chart  XXXVI  (3). 


1074  BUSINESS  CYCLES 

(2)  Brokers'  Loans,  New  York  City.    From  1919  to  1926  the  figures  are  for  (call  and 
time)  street  loans  placed  by  New  York  City  reporting  banks  on  their  own  account,  and  for 
correspondents.     For  some  banks  these  figures  do  not  include  loans  to  dealers  in  securities. 
From  1926  to  1935  the  figures  are  monthly  averages  of  the  loans  to  brokers  and  dealers  on 
security  collateral,  made  by  the  weekly  reporting  member  banks  in  New  York  City.     They 
represent  call  and  time  loans  for  own  account,  for  account  of  out-of-town  banks,  and  for 
account  of  others,  as  compiled  by  the  Federal  Reserve  Board.     The  two  series,  though  very 
imperfectly  comparable,  were  spliced  together.     A  much  better  one  is  available  since 
September  1935. 

(3)  Industrial  Stock  Prices.     This  index  was  taken,  by  permission^  from  the  Standard 
Trade  and  Securities  Service,  Standard  Statistical  Bulletins.     It  is  the  monthly  average 
of  their  weekly  stock  price  index  (351  industrials).     Weighting  is  according  to  the  number 
of  shares  outstanding.     The  average  for  1926  =  100. 

(4)  Bond  Yields.     This  index  was  taken  by  permission  from  the  Standard  Trade  and 
Securities  Service,  Standard  Statistical  Bulletins.     It  is  an  arithmetic  average  of  yields  to 
maturity  of  60  high-grade  bonds,  15  industrials,  15  railroads,  15  public  utilities,  and  15 
municipals.     Monthly  averages  of  Wednesday  closing  prices. 

(5)  Net  Demand  Deposits,  New  York  City.     See  Chart  L  (1). 

(6)  Debits,  New  York  City.     See  Chart  L  (3)  [Outside  Debits]. 

(7)  Railroad  Stock  Prices.     This  index  of  20  railroads,  weighted  by   the  number  of 
shares  of  stock  outstanding,   was  taken  by  permission  from  the  Standard   Trade  and 
Securities  Service,  Standard  Statistical  Bulletins  (e.g.,  see  April  1934,  p.  30).     1926  =  100. 

(8)  Loans  and  Discounts,  New  York  City.     See  Chart  L  (5)   [Outside  Loans  and 
Discounts]. 

CHART LIII.    ISSUES  FOR  "NEW  CAPITAL,"  NEW  YORK  STOCK  EXCHANGE. 

ANNUAL 

Figures  are  for  total  corporate,  domestic  and  foreign,  security  issues  for  "New  Capital" 
by  industrial  groups,  see  Statistical  Abstract  of  the  United  States,  1934,  p.  278.  Preferred 
stocks  of  no  par  value  and  all  common  stocks  are  taken  at  their  offering  price,  other  issues 
at  par.  See  the  Commercial  and  Financial  Chronicle,  Mar.  26,  1921,  p.  1216,  for  a  discussion 
of  the  figures,  of  which  it  is  the  original  source. 

CHART  LIV.     UNITED  STATES.    LOG  SCALE 

(1)  Failures.     Liabilities,  Monthly.     Figures  taken  by  permission  from  the  Standard 
Trade  and  Securities  Service,  Standard  Statistical  Bulletins,  originally  compiled  by  Messrs. 
Ogle,  Dunn  &  Company  and  others. 

(2)  Production  of  Producers'  Goods.     See  Chart  XLIII  (1). 

(3)  "Productive"  Capital  Flotations.     This  series  was  used  with  the  permission  of 
Moody's  Investors'  Service.     Issues  for  refunding,  acquisitions,  working  capital,  and  issues 
of  insurance  companies,  banks,  etc.,  are  excluded,  as  well  as  half  of  the  issues  for  "Real 
Estate  Mortgages"  and  "General  Corporate  Purposes,"  which  are  classed  as  indeterminate. 
They  are  in  any  case  a  very  small  percentage  of  the  total.     Federal  issues  are  excluded, 
but  the  figure  for  October  1934  includes  530  millions  issued  by  the  Home  Owners*  Loan 
Corporation,  because  they  were  for  the  purpose  of  financing  additions  and  improvements. 
According  to  the  terminology  adopted  in  this  book,  the  case  for  including  this  item  is  at 
best  a  weak  one. 

(4)  Corporate  Earnings.     The  figures  of  earnings  of  all  corporations,  i.e.,  receipts  less 
statutory  deductions,  are  from  Statistics  of  Income,  Report  of  the  Commissioner  of  Internal 
Revenue,  Treasury  Department,  and  have  been  corrected  by  W.  L.  Crum  for  dividends 
received  from  other  corporations.     The  author's  thanks  are  due  to  Professor  Crum  for 


APPENDIX  1075 

kindly  permitting  the  use  of  unpublished  material.     The  items  for  1931,  1932,  and  1933, 
being  negative,  could  not  be  plotted. 

(5)  New  Incorporations.  Annual.  Figures  are  for  new  enterprises  with  an  authorized 
capital  of  $100,000  or  more.  They  were  taken  from  the  Statistical  Abstract  of  the  United 
States,  1928,  p.  309.  After  that  year,  this  information  was  discontinued  and  would  have 
to  be  unearthed  in  the  48  statehouses,  an  undertaking  much  beyond  the  means  at  the  dis- 
posal of  the  writer. 

CHART  LV.    GERMANY.    LOG  SCALE 

(1)  Call  Rate.     1924-1935.     This  is  the  monthly  average  of  the  call  money  rate 
(Taggeld),  taken  for  1925  to  1935  from  the  Institut  ftlr  Konjunkturforschung,  Konjunktur- 
Statistisches  Handbuch,  1936,  p.   112.     The  figures  for  1924  are  from  the  International 
Abstract  of  Economic  Statistics,  1919-1930,  p.  93  (International  Conference  of  Economic 
Services). 

(2)  Stock  Price  Index.     1924-1935.     This  is  the  monthly  index  of  the  Statistische 
Reichsamt  (1924-1926  =  100).     It  covers  about  325  stocks,  including  mining  and  heavy 
industries,  manufacturing,  trade  and  transportation,  and  banking.     See  Otto  Donner, 
Die  Kursbildung  am  Aktienmarkt,  Vierteljahrshcfte  zvr  Konjunkturjorschung,  Sonderheft 
36,  1934.     The  Berlin  Stock  Exchange  was  closed  from  July  1931  to  April  1932. 

(3)  Stock  Exchange  Turnover  Tax  Receipts.     In  order  to  get  some  sort  of  index  of 
stock  exchange  transactions,  the  monthly  yield  of  the  stock  exchange  turnover  tax  (taken 
from  the  Institute  fur  Konjunkturforschung,   Konjunktur-Statistischcs  Handbuch,   1936, 
p.  115)  was  adjusted  for  changes  in  the  rate  of  taxation. 

(4)  Industrial  Stock  Issues  (on  the  German  Stock  Exchanges).     Annual  figures  taken 
by  permission  from  Die  Wirtschaftskurve  (a  publication  affiliated  to  the  Frankfurter  Zeitung), 
vol.  XIII,  1934,  p.  267.     The  reference  is  to  compilations  of  the  Statistische  Reichsamt. 
For  1933  the  figures  include  the  Steuergutscheine,  see  Chap.  XV,  sec.  F.     The  1934  figure 
is  for  the  first  10  months. 

(5)  Industrial  Bond  Issues.     See  Series  (4). 

(6)  Dividends.     Per  cent  per  annum.     For  source  and  method  of  calculation,  see 
Chart  XXXVIII  (5). 

(7)  Failures.     Annual.     1925-1934.     Figures  are  from  the  Statistisches  Jahrbuchfiir 
das  Deutsche  Reich,  vol.  LV,  1934,  p.  368,  and  represent  nominal  capital  of  companies 
instituting  liquidation  or  bankruptcy  proceedings  (Einleitung  dcs  Liquidationsverfahrens 
or  Konkurserojfnung).     Of  course,  this  does  not  include  reorganizations  etc.,  which  eco- 
nomically may  be  equivalent  to  failures. 

(8)  Foundations.     Annual.     1925-1935.     Total  share  capital  of  new  enterprises  taken 
from  the  Statistisches  Jahrbuch  fur  das  Deutsche  Reich. 

CHART  LVI.     GREAT  BRITAIN.    LOG  SCALE 

(1)  London  Town  Clearings  (as  Distinguished  from  Metropolitan  and  Country  Clear- 
ings).    Monthly  figures  for  three  weeks  covering  two  stock  exchange  settlement   days, 
consols  settlement  day  and  the  fourth  of  the  following  month.     They  were  taken,  by  per- 
mission, from  the  Monthly  Bulletins  of  the  London  and  Cambridge  Economic  Service. 

(2)  Industrial  Stock  Prices.     The  old  index  (1919-1924)  is  an  arithmetic  average  of  the 
price  relatives  of  20  industrial  ordinary  shares,  (1913  =  100).     The  last  day  of  the  month 
to  December  1922,  and  then  the  middle  of  the  month.     Weighting  is  based  on  1907  Census 
of  Production.     The  new  index  (from  1924  on)  is  on  the  base  1924  =  100,  and  includes 
prices  of  the  shares  of  over  90  companies,  operating  principally  or  exclusively  in  the  United 
Kingdom  (banks  and  railways  are  excluded).     It  is  an  arithmetic  average  of  midmonthly 
figures.     Industrial  groups  are  weighted  according  to  the  1924  Census  net  output  figures. 


1076  BUSINESS  CYCLES 

individual  stocks  according  to  their  market  values  in  1924.  The  new  index  was  adjusted 
to  the  old  one  by  putting  the  1924  average  equal  to  168  (see  Special  Memo  88  of  the  London 
and  Cambridge  Economic  Service,  A  New  Index  of  Prices  of  Securities  by  A.  L.  Bowley, 
G.  L.  Schwarz,  and  K.  C.  Smith,  p.  14).  Both  indices  were  used  with  the  permission  of 
the  London  and  Cambridge  Economic  Service. 

(3)  London  Call  Rate.     Day-to-day  rate.     Averages  for  the  week  ending  the  fifteenth 
of  each  month,  taken,  by  permission,  from  the  London  and  Cambridge  Economic  Service 
Monthly  Bulletins. 

(4)  Money  at  Short  Notice  to  Stock  Exchange.     Data  are  for  10  London  Clearing  Banks 
from  the  Report  of  the  Committee  on  Finance  and  Industry,  1981,  p.  284.     They  do  not 
include  money  at  call  or  money  at  short  notice  to  the  money  market.  *  In  the  case  of  the 
National  Bank,  Ltd.,  only  figures  relating  to  offices  in  England  were  included.     Some  items 
in  the  earlier  years  are  estimates. 

CHART  LVII.    ISSUES  FOR  "NEW  CAPITAL."     GREAT  BRITAIN.    ANNUAL 

New  issues,  domestic  and  foreign  (conversions  excluded),  of  stocks  and  bonds  taken 
from  the  London  Economist,  Annual  Commercial  History  and  Review. 

(1)  Railways.     British,  colonial,  and  foreign.     In  1918  and  1919,  only  colonial  and 
foreign,  and  in  1916  and  1917  only  British  and  foreign. 

(2)  Mines.     Includes  Australasian,  South  African,  and  others. 

(8)  Public  Utilities.     Include  electric  light,  power,  and  telegraph;  tramway  and  omni- 
bus; gas  and  water. 

(4)  Shipping.     Docks,  harbor,  and  shipping. 

(5)  Iron  and  Steel.     Iron,  steel,  coal,  and  engineering. 

(6)  Real  Estate. 

(7)  Motors.     Motor  traction  and  motorcar  manufacturing. 

(8)  Rubber. 

(9)  Oil. 

CHART  LVIII.    UNITED  STATES. 

CONSPECTUS  OF  FEDERAL  RESERVE  BANK  OPERATIONS. 
ARITHMETIC  SCALE 

All  the  series  were  taken  from  the  Annual  Reports  of  the  Federal  Reserve  Board  and  the 
Federal  Reserve  Bulletins. 

(1)  Total  Reserves.     This  includes  nonreserve  cash  reported  separately  after  Decem- 
ber 1922.     Total  reserves  are  called  due  from  the  United  States  Treasury  after  January 
1984.     Gold  included  in  these  accounts  is  valued  at  $20.67  per  fine  ounce  before  January 
1934  and  $35  after  that. 

(2)  United  States  Securities. 

(3)  Bills  Bought. 

(4)  Float.     From  January  1919  to  December  1926  this  series  represents  the  difference 
between  Uncollected  Items  and  Deferred  Availability,  Notes  of  Other  Reserve  Banks  being 
included  in  Deferred  Availability.     From  January  1927  to  March  1936  the  series  repre- 
sents the  sum  of  Float  as  reported  and  Notes  of  Other  Reserve  Banks  (separately  reported 
after  December  1926). 

(5)  Notes  in  Circulation.     The  items  included  in  this  series  are,  from  January  1919  to 
May  1924,  Total  Notes  in  Circulation  and  Federal  Reserve  Bank  Notes  in  Circulation; 
from  June  1924  to  December  1926,  Federal  Reserve  Notes  in  Circulation;  from  January 
1927  to  March  1936,  Federal  Reserve  Notes  Outside  the  Federal  Reserve  Banks,  Federal 
Reserve  Notes  Held  by  Other  Federal  Reserve  Banks,  and  Federal  Reserve  Bank  Notes  in 
Circulation. 


APPENDIX  1077 

(6)  Bills  Discounted. 

(7)  Government  Deposits. 

(8)  Reserve  Account  of  Member  Banks  Calculated — i.e., 

(1)  +  (2)  +  (3)  +  (4)  -  (5)  +  (6)  -  CO. 

(9)  Actual  Reserve  Account  of  Member  Banks, 

CHART  LIX.    UNITED  STATES. 

MAJOR  RESERVE  FACTORS  AND  CENTRAL  MARKET  RATES. 
ARITHMETIC  SCALE 

(A)  Total  Reserves.     See  Chart  LVIII  (1). 

(B)  =  A  plus  United  States  Securities.     See  Chart  LVIII  (2). 

(C)  =  B  plus  Bills  Bought.     See  Chart  LVIII  (3). 

(D)  -  C  plus  Float.     See  Chart  LVIII  (4). 

(E)  =  D  minus  Notes  in  Circulation.     See  Chart  LVIII  (5). 

(F)  =  Bills  Discounted.     See  Chart  LVIII  (6). 

(G)  Rate  of  the  Federal  Reserve  Bank  of  New  York.     From  Federal  Reserve  Bulletin. 
(H)  Commercial  Paper  Rate.     Rate  on  60-  to  90-day  prime  commercial  paper  at  New 

York,  taken,  by  permission,  from  the  Standard  Trade  and  Securities  Service,  Standard 
Statistical  Bulletins. 

(I)  Month-to-month  Changes  in  Total  Reserves. 

(J)  Yield  on  Government  Certificates.     Same  source  as  Series  (H). 

CHART  LX.    BANK  OF  ENGLAND  FIGURES.    LOG  SCALE 

All  figures  are  annual  averages  of  Wednesday  figures,  as  published  in  the  Bankers' 
Almanac  and  Yearbook  and  the  Bank  of  England's  Statistical  Summaries.  On  Nov.  22, 
1928,  the  note  issues  of  the  Bank  of  England  and  of  the  Currency  Note  Commissioners  were 
amalgamated,  and  the  amount  of  the  fiduciary  note  issue  was  fixed  at  260  million  pounds. 
Therefore,  figures  of  reserves  and  securities  are  not  comparable.  On  Aug.  5,  1931,  the 
maximum  of  the  fiduciary  note  issue  was  increased  to  275  million  pounds,  and  in  April 
1933  it  was  put  back  to  260  million  pounds.  No  attempt  has  been  made  to  present  figures 
about  notes,  reserves,  or  securities  for  1928.  From  1929  to  1934  Other  Securities  equal 
Discounts  and  Advances  plus  Securities. 


Index 


AAA  legislation,  988-992 

Abbott,  C.  C.,  813n. 

Abel-Musgrave,  R.,  1037n. 

Acceleration,    principle    of,    181-182,    916, 

925 

Accumulation,  75—84 
definition  of,  75 
original,  229-230 

(See  also  Corporate  Accumulations) 
Acquisitive  principle,  495 
Adaptation,  waves  of,  180,  183 
Addison  Act,  752 
Advances  of  London  clearing  banks,  1931— 

1937,  959 

Agents  de  change,  679 
Aggregate    net    income    of    corporations, 

United  States,  postwar,  831-832 
Aggregative    theories    of    business    cycles, 

inadequacy  of,  144 

Agrarian  depressions,  in  1820's,  267-270 
in   last    quarter    of   nineteenth   century, 

320-325 

in  1920's,  732-743 
Agricultural  developments,   237-240,   266- 

270,  319-325,  401-402,  732,  743 
Agricultural  machine  industry,  American, 

317n.,  389 

Agricultural  overproduction,  735n. 
Akerman,  Johan,  182 
Alford,  L.  P.,  and  Hannum,  J.  E.,  802n. 
Alfter,  J.,  95n. 
All  other  accounts,  891 
Allen,  G.  C.,  374n. 


Altschul,  E.,  and  Strauss,  F.,  738n. 
American  Telephone  and  Telegraph  Com- 
pany, index  of  general  business,  23n., 

Amoroso,  Luigi,  46w.,  185 

Anderson,  B.,  852n.,  854«.,  859w. 

Anderson,  O.,  193n.,  202n. 

Andrew,  A.  P.,  664n. 

Angell,  J.,  598n.,  886n. 

Animal  cycles,  530 

Annalist,  26 n. 

Anticipations,  53-55,  140 

Antisaving  attitude,  699-700 

Applegate,  LaRue,  1025 

Armaments,  965 

Armstrong,  C.  E.,  165 

Arthur,  H.  B.,  459w. 

Australian  mining  boom  (see  South  African) 

Austro-German    customs    union,    plan    of, 

930-931 

Autodeflation,  136 

Automobile  industry,  rise  of,  415-417 
Autumnal  drain,  656 
Axe-Flinn  index  of  business  conditions  for 

Germany,  23  n. 
Axe-Houghton  index  of  business  activity, 

24n. 

Ayres,  Colonel,  12n.,  166n.,  490n. 
Ayres,  M.  V.,  826 


B 


Bagehot,  W.,  677n. 
Baker,  E.  F.,  785 

Balance  of  payments,   American  postwar, 
707-708 


*  This  Index  does  not  contain  references  to  the  numerous  industries,  concerns,  and 
individual  entrepreneurs  that  have  been  mentioned  in  this  book,  especially  in  Chapters 
VI,  VII,  XIV,  and  XV.  Items  that  occur  only  in  the  Appendix  have  also  been  excluded* 

1079 


1080 


BUSINESS  CYCLES 


Balance  sheet  of  Federal  Reserve  banks, 

combined,  888-894 
Balances,  demand  for,  603 

supply  of,  606 
Balfour  Committee,  29 n. 
Balogh,  T.,  680n. 
Bank,  assets,  quality  of,  116n. 
clearings,  16 
credit  in  national  socialist  Germany,  981— 

982 

debits,  16 

epidemics,  912,  938,  943-944 
failures,  16 
Prussian,  308 
rate,  16 

in  England,  1919-1929,  904-905 

(See  also  Reserve  bank  rate) 
rates  in  1931,  933 
reform  of  1908,  German,  400-401 
reforms,  nature  of,  661 
reserves,    economy   of,    due   to   Federal 

Reserve  Act,  885n. 

(See    also    Investments;    Loans;    Sec- 
ondary reserves) 
Bank  of  England,  postwar  policy  of,  904- 

905 

rate  of,  625-626 

variations  in  gold  stock  of,  676-677 
Bank  of  North  America  (1782),  293 
Bank  of  United  States,  failure  of,  911 
Bank  Act,  of  1844,  307 
of  1933,  987 
of  1935,  1027 
Emergency,  987 
Bankers'  acceptances,  624,  625 
Bankers'  banks  (see  Banks) 
Bankers'  loans  (see  Loans) 
Bankhead  Act,  991 
Banking,  commercial  theory  of,  115 
investment  theory  of,  115,  862 
reckless,  260 
wildcat,  299-300 
Banking  series,  postwar,  850-870 
Banks,  bankers',  112 
central,  112 

and  crises,  660,  661 
policy  of,  648-665 
and  stock  speculation,  689-691 
Federal   Reserve   (sec  Federal   Reserve 
Banks) 


Banks,  industrial  or  promoting  in  Germany, 

348-350 

land  (see  Landschaften) 
member,  112 

and  Federal  Reserve  Board,  867-868 
(See  also  Investments;  Loans;   Re- 
serves) 
mortgage,  German  (Hypothekenbanken), 

359,  618 

national,  310,  315,  316 
special,  in  Germany,  406 
of  United  States,  292 

(See  also  Reichsbank) 
Baring  crisis,  381-382 
Barnett,  G.  E.,  418n. 
Barns,  A.  F.,  520n. 
Barrett,  D.  C.,  315n. 
Barter  terms  of  trade,  667 
Baude,  W.  A.,  215n. 
Bauer,  O.,  696n. 
Beach,  W.  E.,  677n. 
Beales,  H.  L.,  353n. 
Bean,  L.,  528,  532n.,  823w. 
Bellman,  Sir  H.,  964n. 
Below,  G.  v.,  229n. 
Belshaw,  H.,  321 
Beneduce  report,  718n. 
Benner,  S.,  532n. 
Bennett,  M.  K.,  237n. 
Bennett  R.,  56n. 
Bercaw,  L.  O.,  520n. 
Berry,  T.  S.,  458ra. 
Bertrand,  J.,  60n. 
Beveridge,  Sir  William,  166n.,  179n.,  458n., 

510n.,  969n. 

Bill  brokers,  English,  625n. 
Bills  drawn,  Germany,  869 
Bills  of  new  tenor,  253n. 

(See  also  Commercial;  Finance; 'Treas- 
ury) 

Birck,  L.  V.,  525n. 
Bismarck's  railroad  policy,  346 
Bituminous  Coal  Conservation  Act,  933n. 
Black,  John  D.,  941n.,  942n.,  989n.,  991n., 

992n. 

Black  Friday,  316,  336 
Bland  Act,  318 
Bliss,  C.  A.,  926n -928n. 
Blodgett,  R.  H.,  524n. 
Bobroff's  moving  correlation,  202n. 
Boehm-Bawerk,   E.  von,   78n.,   223,   505, 

548,  561,  836n. 


INDEX 


1081 


Bond  prices,  Moody 's  series  of,  61 9n. 
Bond  yield,  16 

behavior  of,  618-622 

index  of  (Standard  Statistics  Co.),  619n. 
Bonds,  collateral  trust,  406 
Borel,  fimile,  194 

Borrowing  (see  Government;  Household) 
Bortkiewicz,  L.  von,  452 
Bouniatian,  M.,  220n. 
Bowley,  A.  L.,  59,  461n.,  462n.,  510n.,  566, 

729n.,  751n.,  799n.,  847,  919,  967n. 
Bowman,  R.  T.,  830n. 
Bradbury  report,  723n. 
Breaks  in  trend,  202 
Bresciani-Turroni,    C.,    78n.,    181n.,    205, 

877n.,  982n. 

Brokers  (see  Bill  brokers;  Stock  brokers) 
Brokers'  loans,  16 
Brown,  E.  C.,  538n. 
Brussels  Conference,  824 
Bryce,  R.,  456n. 
Bubble  Act,  248 
Building,  ahead  of  demand,  158n. 

contracts  awarded,  16,  19 

permits,  16,  19 

residential,  19,  159 

societies,  English,  753 
Building  boom,  German  (1872),  359 

great  English,  964-965 
Building  booms  of  1920's,  743-753 
Bullock,  C.  J.,  181,  260n.,  667n.,  987n. 
Billow  duties,  401 
Burgess,  R.,  888n. 

Burns,  A.  F.,  22n.,  497n.,  498n.,  797n. 
Burr,  Ebersole,  and  Peterson,  831n. 
Business,  activity,  16 

cycle  theory  (see  Aggregative  theories) 

indices,  12n.,  23,  24 

normal,  4,  5 

situations,  3 

volume  of,  16 
Business  Annals,  lln.,  221 
Businessman's  normal,  4 
Buy-now  campaigns,  587 


Cairncross,  F.,  590n. 

Call  rate,  624 

Call  rates,  England,  882-884 

Germany,  880-881 

New  York,  876 


Campbell-Bannerman,  Sir  H.,  399,  727n. 
Capacity  (see  Excess) 
Capital,  42,  129 

export,  American,  after  World  War,  703, 

707 
British,  367,  381 

gains  tax,  1040 

goods  (see  Interest) 

import  German,  postwar,  718-722 

issues  (see  Issues) 

market,  129 
Capitalism,  competitive,  96 

definition  of,  223 

rise  of,  228 

stationary,  1033 

trustified,  96,  145 

Capitalist  and  entrepreneur,  103-104 
Caprivi  treaties,  308 
Carli,  G.  R.,  451 
Cartel  quota,  62 
Cash  held  by  corporations,  860n. 
Cassel,  Gustaf,  38,  472,  680n. 
Causal  factors,  25 
Central  banks  (see  Banks) 
Central  creation  (see  Credit) 
Central  market,  126 

American,  theory  of,  892-894 
Certified  check  accounts,  680 
Chain  method,  456 
Chain  stores,  sales  of,  16,  19 
Chamberlain  Act,  752,  918 
Chamberlin,  E.  H.,  56n.,  63,  65n.,  67n. 
Champernowne,  D.  G.,  80n.,  968n. 
Change,  autonomous,  14 

conditioning  of,  72 

external  factors  of,  6-13 

internal  factors  of,  72-87 
Changes,  in  consumers'  tastes,  73-74 

structural,  lln.,  12n. 
Changing  harmonics,  18 In. 
Chapman,  J.  M.,  650n. 
Chawner,  L.  J.,  476n.,  745n. 
Chinese  indemnity  to  Japan,  677 
Circulation,  velocity  of,  545-546 

(See  also  Currency) 
Civil  War,  334 

deflation  after,  315-317 

economic  effects  of,  314-315 
Clapham,  J.  H.,   252n.,   254,  274n.,   276, 

303n.,  367n.,  565n. 

Clark,  Colin,  729n.,  765n.,  798,  799,  828- 
829,  849n.,  850,  885,  97 On. 


1082 


BUSINESS  CYCLES 


Clark,  G.  N.,  233n. 

Clark,  J.  B.,  64 

Clark,  J.  M.,  560n.,  806n. 

Clark,  V.  S.,  231n.,  252n.,  290,  384,  409n., 

415n.,  421n.,  565 
Class  consciousness,  699n. 
Clausing,  G.,  859n.,  766 
Clean  cyclical  series,  200 
Clean  trend  series,  200 
Clearing-house  certificates,  334w. 
Cleveland  Trust  Company,  business  index 

of,  12n.,  23n. 
Coal  famine,  413 
Coal  mining,  242 
Coats,  R.  H.,  460n. 
Cobb  and  Douglas,  590rc. 
Cobweb  problem,  49 
Coffee  cycles,  530 

Cole,  A.  H.,  252n.,  392n.,  471,  562n.,  582w. 
Collateral  loans,  18,  583 
Colm  and  Lehmann,  1002n.,  1039n. 
Colonial  Stock  Act  (1900),  430n. 
Columbia  Committee  on  Economic  Recon- 
struction, 802n. 

Commercial  and  Financial  Chronicle,  26  n. 
Commercial  bills,  15 

decreasing  importance  of,  625 
Commercial  crisis,  257 
Commodity  credit  corporation,  991 
Commodity  stocks,  524 
Communist  Manifesto,  497 
Companies  Registration,  Act  of  1844,  307 
Competition,  cutthroat,  61 

imperfect,  56-68 

monopolistic,  62,  63-68 

perfect,  46 

wastes  of,  60 
Condliffe,  J.  B.,  924n. 
Conklin,  W.  D.,  476n. 
Consequential  factors  (series),  25 
Consumers'  credit  (see  Public  spending) 
Consumers'  expenditure,  236,  299 

and    realized   income,    postwar,    United 

States,  823-827 

Consumers'  prosperity,  718,  721,  811 
Consumption,  15 

Continental  paper  currency,  253n. 
Continental  system,  257 
Copeland,  M.  T.,  393n.,  819,  820n. 
Corporate  accumulations,  United  Kingdom, 
postwar,  829 

United  States,  postwar,  820-822 


Corporate  state,  697 

Corporate  surplus  and  real  investment,  822 
Corporation  tax  (see  Tax) 
Corsten,  H.,  222n. 
Cost,  decreasing,  law  of,  90 
increasing,  law  of,  90 
of  living  (see  Price  level) 
Cost  curves,  89-91 
Costs,  index  of,  476 
Cotton  mania,  288          % 
Cotton  program,  effects  of,  991-992 
Cournot,  A.,  59,  60n. 
Cox,  G.  V.,  22n. 
Crandall,  R.,  458n.,  470n. 
Credit,  absorption  of,  by  stock  speculation, 

680-681 

central  creation,  650 
autonomous,  893 
responsive,  893 
member  creation,  650 
rationing  of,  651 

(See    also    Bank   credit;    Commodity 

credit  corporation) 
Credit  creation,  meaning  of,  110-114 

and  railroads,  329 
Credit  mobilier,  347n. 
Crises,  162 

English,  in  seventeentn  and  eighteenth 

centuries,  249-250,  296,  297 
no  technical  meaning  of,  5 
theories  of,  162 

(See  also  Baring,  Commercial,  and 
World  crises;  Central  banks;  Waren- 
handels  Krisen) 

Crisis  of  concerns,  in  Germany,  763 
"Crisis  of  the  dollar,"  937 
Crisis  of  1810-1811,  262 
of  1815,  262,  298 
of  1818,  American,  301 
of  1825,  279,  299 
of  1836-1839,  299-300 
of  1847,  in  England,  344-345 

in  Germany,  350 
of  1857,  331 
in  England,  377 
in  Germany,  361-362 
of  1866,  in  England,  378 
of  1873,  336-337 
in  England,  380 
in  Germany,  362-364 
of  1883-1884,  340 


INDEX 


1083 


Crisis  of  189S,  341 

in  America,  388-389,  402,  403 
in  England,  382 
in  Germany,  366 
of  1896,  317 
of  1907,  in  Germany,  448 

in  United  States,  407,  408,  410,  424- 

429 

of  1921,  in  United  States,  786 
Crops,  variations  of,  8,  176-179 
Croston,  J.  J.,  and  Davenport,  D.  H.,  841n. 
Crum,  W.  L.,  20,  165,  166n.,  169n.,  181, 
210rz.,  212w.,  498w.,  557w.,  615n.,  629, 
814,    819,    830n.,    832,    833/t.,    924n., 
1006n. 

Crum  and  Patton,  193n. 
Cumulation,  181 
Cunliffe  report,  723 
Cunningham,  W.,  254n. 
Currencies,  gold,  postwar,  703 
Currency,     in     circulation     and     deposits, 

ratio  of,  886-887 
Currency  school,  640 

Current  accounts,  postwar  behavior  of,  870 
Currie,     L.,     851n.-853n.,     855n.,     856«., 
859n.,    900n.,    996n.,    lOOlrc.,    1002«., 
1012n. 

Cycle,  formal  definition  of,  200 
Cycles,  22,  138 

multiplicity  of,  161 

(See  also  Animal;  Coffee;  Four-phase; 
Harvest;     Hog;     Juglar;     Kitchin; 
Kondratieff;    Special;   Specific;   and 
Shipbuilding  cycles 
Cyclical  industries,  158 
Cyriaci-Wantrup,  S.  von,  618 


D 


Dailey,  D.  M.,  854n. 

Darmois,  G.,  193n. 

d'Avenel,  Vicomte,  458w. 

Davenport,  D.  H.,  and  Croston,  J.  J.,  841n. 

Davis,  J.  S.,  989n.,  991n. 

Dawes  loan,  715 

Dawes  plan,  702-704 

Day,  E.  E.,  490«. 

Debt-deflation,  146,  909,  925 

Debits,  1014 

and  deposits,  postwar  relation  of,  857-858 
and  loans,  postwar  relation  of,  858-859 
postwar  behavior  of,  817-820 


Deferred  availability,  890n. 

Deflating  of  monetary  expressions,  456 

Deflation,  260-266 

after  Civil  War,  315-317 

in  England,  postwar,  725 

in  Germany,  in  1930,  923 

in  1931,  935 

Deflation  and  inflation,  260 
Deflationary  effect  of  saving,  587 
Deflationary     effects     of     social     security 

payments,  1032n. 
Deflationary  pressure,  939-940 
Deflection,  579 
Demand,  for  durable  goods  (see  Interest) 

for  money,  548 

replacement  of,  158,  189-191 
Demand  deposits  (see  Deposits) 
Demonetization  of  silver,  in  Germany, 

308-309 

Department  stores,  sales  of,  16 
Deposits,  16,  950,  1007 

compensated,  120 

created,  113 

and  currency  (see  Currency) 

and  debits  (see  Debits) 

demand,  adjusted,  851 
net,  851 

and  loans  (see  Loans) 

in  London  clearing  banks,  1931-1937,  959 

original,  119 

outside,  payroll  covariation  with,  573  n. 

owned,  119 

time,  583 

nature  of,  850-857 

velocity  of,  579,  585,  856 
Depression,  149 

deep,  154 

Great,  1873-1896,  353-354 

(See     also     Agricultural     and     Agrarian 

depressions) 

Depressionless  industries,  1010 
Descriptive  trends,  201 

Determinateness  of  economic  quantities,  41 
Devaluation,  American,  998 

and  stock  prices,  1000 
Devons,  E.,  799n. 
Dewey,  D.  R.,  315w. 
Dieterici,  K.  F.  W.,  252n. 
Discount  houses,  625n. 
Discounting,  new  system  of,  625 

process  of,  609 

of  quasi-rent,  61  In. 


1084 


BUSINESS  CYCLES 


Discounts,  16 

and  loans  (see  Loans) 
Discoveries,  8-9 
Disequilibrium,  47 
Disorganized  markets,  540 
Dispersion  graphs,  521  n. 
Disproportionalities   in   price   movements, 

948-950 

Dissaving    and    gains    from    stock    trans- 
actions, 679 
Dividends,  17 

and  earnings  in  1931,  951-952 
Divisia,  Francois,  452 
Dodge  Corporation,  745n. 
Donner,  Otto,  21  n. 

Douglas,  P.  H.,  510n.,  565,  798, 799n.,  88671. 
Douglas  and  Cobb,  590n. 
Dumping,  61 

Dunbar,  Charles  F.,  220n.,  830,  S34n. 
Dunlop,  J.,  277n. 
Duopoly,  60-61 
Dutot,  451 

Dyer  and  Martin,  894n. 
Dynamics,  48 


E 


Earnings  and  dividends  (see  Dividends) 
Easy  money,  17 

(See  also  Stock  pricing) 
Easy  money  policy,  658 
Ebersole,  Burr,  and  Peterson,  831n. 
Economic  evolution,  86 
Economic  horizon,  40,  99 
Economic  policies  in  America,  1981-1932, 

939-943 

Economic  system,  41 
Economist's    Index    of    Business    Activity, 

24n. 

Econostat,  business  index  of,  23n.,  24n. 
Eddy,  G.  A.,  879n. 

Edgeworth,  F.  Y.,  50,  60n.,  62,  452,  459n. 
Edie  and  Weaver,  585n. 
Edison  Electric  Institute,  411 
Efficiency  of  money,  546 
Effort  elasticity,  159n.,  177,  324 
Ehrenberg,  V.,  225n. 
Ehrke,  K.,  476n. 
Eiteman,  W.  J.,  872n. 
Elastic  waves,  179-183 
Elasticity,  product,  38n. 
of  substitution,  80 


Electricity  Supply  Act,  of  1919,  43S 

of  1926,  758 

Emergency  Banking  Act,  987 
Emergency  Farm  Relief  Act,  1933,  987,  990 
Emergency  Relief  Act,  1933,  1001 
Employment,  16,  509-519,  1009,  1011,  1015 

Germany  and  United  States,   1931  and 
1932,  945-946 

in  national  socialist  Germany,  971 
Enclosures,  238 
Endogenous  factors,  7n. 

theories,  7n.,  48 
England,  1931-1938,  954-971 
Entrepreneur,  enterprise,  102-104 
Entrepreneur  and  capitalist,  103-104 
Epstein,  R.  C.,  415n.,  416,  773n.,  830n., 

832».,  833n.,  834 
Equilibrium,  aggregative,  48 

concept  of,  80-71 

general  or  Walrasian,  42 

imperfect,  44 

neighborhood  of,  71,  173 

partial  or  Marshallian,  43 

perfect,  44 

sloppy,  44 

state  of  42 

use  of,  68-71 

zone  of,  58 

Equilibrium  values,  42 
Ernie,  Lord,  267n. 
Erratic  series,  19 
Errors  of  judgment,  role  of,  140 
Essars,  P.  Des,  586 
Estate  tax  (see  Tax) 
Evans,  G.  C.,  57n.,  184n.,  528 
Evolution,  economic,  86 
Excess  capacity,  54,  61,  62,  66,  67,  509,  754, 
802-808 

reserves  (see    Reserves) 
Exchange,  equation  of,  44 

foreign  (see  Foreign  exchange) 
Exchange  equalization  account,  957n. 
Exogenous  factors,  7n. 
Expectations,  53-55,  140 
Expenditure    of    households,    relation    to 
household's  receipts,  549n.,  550n. 

and  relief  in  1931-1932,  942-945 

(See  also  Consumers',  Federal,  System 

expenditures) 

Exploitation,  theory  of,  229 
Exports,  15 

British,  367 


INDEX 


1085 


Exports,  British,  postwar,  755-756 

postwar,  from  England  to  Germany,  703 
External  economies,  92,  239,  875,  390 
External  factors,  6-13,  700 
Ezekiel,    M.,   520n,,    522n.,    532n.,   736n., 
819n.,  951n. 


Fabricant,  S.,  821,  831n.,  833,  913n.,  950n. 

Failures,  16 

Farm  Credit  Act,  988n. 

Farm  Mortgage  Corporation  Act,  987 

Farming  situation  in  1931  and  1932,  941- 
942 

Fascist  state,  697 

Feavearyear,  A.  E.,  829 

Federal  expenditure,  postwar,  709 

Federal  reserve  banks,  combined  balance 
sheet  of,  888-894 

Federal  Reserve  Board,  and  member  banks, 
867-868 

Federal  Reserve  System,  assistance  ren- 
dered to  European  Central  Banks, 
888n. 

mechanics  of,  888-894 
policy  of,  894-904 

and  causation  of  world  crisis,  902 

Fetter,  F.  A.,  497 

Finance  bills,  624 

Financing,  primitive,  230 

Finanzkapitalismus,  405n. 

Fiscal  policy,  in  bourgeois  era,  310-311 
postwar,  709-714,  716-722,  729-731 
in  United  States  since  1932,  1038-1042 

Fisher,  Irving,  35n.,  127n.,  129n.,  143n., 
146-147,  185,  212n.,  452,  455n.,  457, 
459n.,  470,  490n.,  497,  509n.,  522n., 
584n.,  746w.,  900n.,  909w.,  938 

Fisher,  R.  A.,  193n.,  202n. 

Fisher's  ideal  formula,  456 

Flight  from  the  land,  325 

Float,  890 

Flux,  A.  W.,  799n. 

Foenus  nauticum,  615 

Folio w-the-leader  system.  61 

Forced  savings,  112n. 

Fordney-McCumber  Act,  706 

Forecasting,  13 

Foreign  exchange,  17 

Foreign  exchanges  and  differences  in  money 
rates,  674 


Foreign  invesments  (see  Investments) 

Foresight,  52 

Forethought,  52 

Forty-months  cycle,  165 

Foster,  R.  R.,  745n. 

Foundations  of  firms,  17,  94 

Fourier  analysis,  applicability  of,  to  time 

series,  198-199 
Four-phase    cycles,    149 
Franklin,  B.,  287 
Fraser,  L.  M.,  877n. 
Free  gold,  887-888 

trade,  in  England,  306-307 
Freehand  methods  of  analysis,  21  In. 
Freight-car  loadings,  17 
French,  D.  R.,  854n. 
French  indemnity,  1871,  313-314 
Frickey,  Edwin,  22n.,  23n.,  200n.,  202n.t 

203,  204,  316n.,  551n.,  615n.,  877n. 
Friction,  42n.,  50 
Frisch,  Ragnar,  48,  60n.,  151n.,  171n.,  181. 

184n.,  189,  195,  208-212,  215-219,  452, 

469,  560n. 
Fullarton,  640 


Gay,  E.,  224n.,  238 

Gayer,  A.  D.,  916n.,  942 

Gehlhoff,  W.,  462n. 

Georgescu-Roegen,  N.  S.,  215 

Gibson  paradox,  633n. 

Giffen,  R.,  590n. 

Gilboy,  E.  W.,  469n.,  527n.,  529n.,  564n., 

565 

Gilfillan,  S.  C.,  85n.,  227n. 
Gill,  C.,  745n.,  749n. 
Gladstonian  finance,  spirit  of,  311 
Glass-Steagall  Act,  939-940 
Goether,  E.  T.,  538n. 
Gold,  free,  887-888 
Gold  account,  inactive,  1028 
Gold-buying  policy,  998 
Gold  currencies,  postwar,  703 
Gold  devices,  676 

discoveries,  8 

flows,  17 

influx  into  United  States,  863n,  887-888, 
997,  1028 

market,  London,  675-676 

and  price  level,  472-473 

production,  variations  in,  176 


1086 


BUSINESS  CYCLES 


Gold  Reserve  Act,  997-998 

scare,  667n. 

South  African,  676 

Standard  Act,  317 

of  1925,  726 

Gold  standard,  England's  abandonment  of, 
955-958 

postwar  working  of,  887-888 
Gold  Standard  Suspension  Act,  955 
Gold  sterilization,  888n.,  1027 
Golden weiser,  E.  A.,  71  In. 
Goodrich,  C.,  805n. 
Gordon,  R.,  888n. 
Goschen,  Lord,  382,  607,  675 
Government  borrowing,   cyclical  behavior 

of,  607 
Government  spending  in  national  socialist 

Germany,  975-977 

Gray,  T.  H.,  and  Terborgh,  G.  W.,  746n. 
Greenstein,  B.,  166n.,  171n. 
Gresham,  Sir  Thomas,  234n. 
Growth,  83-84,  158 

(See  also  Organic  growth) 
Gruenbaum,  F.,  550w. 
Guyot,  Yves,  15 


H 


Haberler,  Gottfried  von,  452,  454n. 

Hahn,  A.,  582n. 

Hakluyt,  R.,  246 

Haldane,  J.  B.  S.,  184n. 

Hall,  L.  W.,  583n. 

Hamilton,  Alexander,  287,  288 

Hamilton,  E.  J.,  225n.,  231,  252n.,  459n., 

564n. 

Hammond,  Bray,  260n.,  294n. 
Hanau,  A.,  532 
Hanch,  C.  C.,  826n. 

Hannum,  J.  E.,  and  Alford,  L.  P.,  802n. 
Hansen,   A.   H.,   78w.,   5COn.,   565,   634n., 

907?i.,  1033n. 

Hardy,  C.  O.,  601  n.,  SSSn. 
Hardy,  C.  O.,  and  Owens,  R.  N.,  680w. 
Harmonics,  changing,  181  n. 
Harms,  Bernhard,  lln. 
Harris,   S.  E.,   758«.,  866n.,   888n.,   958n., 

961n.,  996n. 
Harrod,  R.  F.,  43n.,  60n.,  64n.,  67n.,  68, 

144,  954w. 

Harvard  method,  21-24 
Harvest  cycle,  176-179 


Hatry  failure,  908 

Hawley-Smoot  Act,  707 

Hawley-Smoot  tariff,  915 

Hawtrey,  R.,  142,  634n.,  724n.,  904n.,  918n., 

956n. 
Hayek,  F.  A.  von,  78n.,  296n.,  333,  345n., 

603n.,  634 

Hayek  effect,  812,  814 
Heckscher,  E.  F.,  245n. 
Heichelheim,  F.,  226n.     % 
Hepburn,  A.  B.,  583n. 
Hesitations,  183,  428 
Hicks,  J.  R.,  56n.,  60n.,  65n.,  80n.,  574n. 
High-pressure  engine,  289 
Hilferding,  R.,  405n.,  696n.,  715n. 
Hirsch,  J.,  760 
Historical  variables,  194-197 
Hoarding,  76,  579 
Hobson,  C.  K.,  367n.,  429 
Hoffmann,  W.,  490n. 
Hog  cycle,  49,  54,  531-533 
Holmes,  C.  L.,  739n. 
Home  Owners'  Loan  Acts,  987 
Hooker,  L.,  572n. 

Hoover,  E.  M.,  391n.,  786n.,  788n. 
Horizon,  economic,  40,  99 
Hoskins,  L.,  244n. 
Hotelling,  H.,  63n. 

Households'  borrowing,  postwar,  826-827 
Housing  Act  of  1936,  1025 
Hubbard,  J.  B.,  945n. 
Hudson,  railway  king,  344 
Hunt,  B.  C.,  278n.,  279n. 
Hutchinson,  W.  T.,  389n. 
Hyndman,  H.  M.,  162n. 
Hypothekenbanken  (see  Banks) 
Hytten,  J.,  147n. 


Illinois  Central  Railroad,  330-331 
Imperialism,  432 

Neo-Marxist  theory  of,  405n. 

sociology  of,  696n. 
Imports,  15 

Inactive  gold  account,  1028 
Inactive  systems,  42n.,  160 
Income,    aggregate    net,    of    corporations, 
United  States,  postwar,  831-832 

generation    of,    by    public    expenditure, 
920n.,  923,  1001-1006 

national,  561,  950,  1001-1002,  1014 


INDEX 


1087 


Income,  parities,  736 

postwar  behavior  of,  in  Germany,  827 
in  United  Kingdom,  828 
in  United  States,  819 

realized  (see  Consumers'  expenditure) 

tax  (see  Tax) 

velocity  of  money,  545 
Incomes,  sum  total  of  money,  15 
Independent  Treasury  Episode,  293n. 
Indeterminateness,  50 
India,  native  capitalism  in,  705 
Indices   of  business  conditions,   American 
Telephone  and  Telegraph   Company, 
23n.,  24n. 

Axe-Flinn,  23n. 

Axe-Houghton,  24n. 

Cleveland  Trust  Company,  12n.,  23«. 

Economist,  The,  24n. 

Econostat,  23n.,  24n. 

New  York  Times  Weekly,  23n. 

Ogburn-Thomas,  24w. 
Individual  series,  25 
Industrial  Advances  Act,  987n. 
Industrial  policies  in  United  States  after 

1933,  effects  of,  1043-1044 
Industrial  revolution,  168,  170,  253 

New,  397 

of  1920's,  753-794 
Inflation,  and  deflation,  260 

safeguards  against,  997n. 
Inflation  Act,  997-998 

Inflationist  interests  and  government,  986w. 
Innovation,  84-86 

resistance  to,  examples  of,  243,  244 

theory  of,  87-192 
Instalment  sales,  826 
Interest,  27,  32,  41,  123-126 

adapted  rate  of,  603,  605,  636 

and  demand  for  durable  goods,  609 

equilibrating  rate  of,  603w.,  605 

equilibrium  rate  of,  603 

lag  in,  605 

natural  rate  of,  126 

postwar  behavior  of,  811-817 

and  quasi-rent,  608 

rate  of,  602-638 

real  rate  of,  126 

and  value  of  capital  goods,  609 
Interest  rates,  and  issues,  880 

on  mortgages,  617-618 

structure  of,  611 
Intermediate  products,  38 


Internal  economies,  92 

Internal  factors,  7n. 

International  relations,  cyclical  aspects  of, 

666-678 

Inventions,  8-9,  84-86 
Investment,  76 

in  England,  postwar,  885 
real,  76 

and  corporate  surplus,  822 
issues  for,  879-880 
temporary,  75n.,  77,  580,  606,  623,  644, 

870-872 

Investment  opportunity,  theory  of  vanish- 
ing, 1032-1036 

Investment  survey,  Moody's,  880 
Investment  theory  of  banking,  115,  862 
Investments,  foreign,  English,  behavior  in 

world  crisis,  962-963 
and  loans,  of  banks  outside  New  York 

City,  645-646 

member  banks',  595,  598,  862-868,  870 
Irregularities,  external,  7n.,  144 

internal,  144 

Issues,  capital,  English  postwar,  883-885 
and  interest  rates,  880 
for  New  Capital,  United  States,  postwar, 

877-879 

for  real  investment,  879-880 
of  securities,  15 

in  national  socialistic  Germany,  983 


Japan,  industrialization  of,  705 

Jenks,  J.,  367n. 

Jerome,  H.,  402,  419,  747n.,  783n. 

Jevons,  W.  St.,  176,  371,  452,  461n.,  469n., 

627n.,  656w. 

Joint  Stock  Companies  Act,  307 
Jones,  G.  T.,  476w. 
Juglar,  Clement,  139,  162,  163,  326n. 
Juglar  cycles  (Juglars),  169 

K 

Kahn,  R.  F.,  80n.,  92,  515,  752n.,  758 
Kaldor,  N.,  50n. 
Kalecki,  M.,  185-188 
Karsten,  K.  G.,  181,  877n. 
Kemmerer,  E.  W.  A.,  490n.,  584 
Kerr-Smith  Act,  990 
Keyes,  E.  W.,  583n. 


1088 


BUSINESS  CYCLES 


Keynes,  J.  M.,  17,  43,  127n.,  585n.,  608n., 

633n.,  832n.,  838n.,  951». 
Kindersley,  Sir  R.  M.,  962n. 
King,  Gregory,  237 
King,  W.  J.,  490n.,  562n.,  820n.,  823 
King,  W.  T.  C.,  292,  378n.,  626n. 
Kitchin,  Joseph,  165,  169n.,  215,  473,  555n. 
Kitchin  cycles  (Kitchins),  169 
Knickerbocker  failure,  428 
Knight,  F.  H.,  52 

Kondratieff,  N.  D.,  164,  470n.,  477n. 
Kondratieff  cycles  (Kondratieffs),  169 
Konsum-Konjunktur,  718 
Kreditanstalt,    Austrian,    breakdown    of, 

980n. 

Kreps,  Thomas  J.,  809n. 
Kreuger  failure,  908 
Kuczynski,  R.,  565ra. 
Kuznets,    Simon,    21n.,    165,    389,    398n., 

497n.,  500n.,  520n.,  801,  819,  823,  912, 

950n.,  951,  lOOln.,  1002n. 


Labor  policies  in  United  States  after  1933, 

effects  of,  104&-1043 

Labor  Relations  Act,  national,  1042,  1045 
Lag,  of  maximum  correlation,  502 

and  sequence,  57n. 
Lag  distribution,  method  of,  509re. 
Lags,  technological,  48 
Land  grants  to  railroads,  328 
Land  sales,  receipts  from,  326 
Land  settlement,  policy  of,  319 
Landschaften  (landbanks),  618 

Prussian,  268n. 
Lange,  Oskar,  49n.,  520n. 
Laspeyres'  formula,  456,  490 
Lausanne  agreement,  936 
Lavington,  F.,  54n.,  430n. 
Law,  John,  250-252 
Layton,  Sir  W.,  461n. 
Layton  report,  718n.,  721 
LeChatelier,  principle  of,  47 
LeCorbeiller,  C.,  210n. 
Lederer,  W.,  827n. 
Lehmann  and  Colm,  1002ra.,  1039n. 
Leonard,  R.,  237n. 
Lenoir,  M.,  501n.,  520n. 
Leonard,  £.,  490n. 
Leontief,  W.  W.,  49n.,   60n.,  452,  476n., 

520n.,  521n.,  575n.,  822 


Leong,  Y.,  490n.,  584n.,  796n. 

Lerner,  A.  P.,  63w. 

Less-than-carload-lots,  17 

Levin,  Moulton,  and  Warburton,  820n. 

Lidderdale,  W.,  626n. 

Limited  Liability  Act,  307 

Lippmann,  W.,  916n. 

Liquidation,  abnormal,  149 

Liquidations,  15 

List,  Friedrich,  115 

Loan,  Dawes,  715 

Loans,  on  account  of  others,  872-874 
all  other,  postwar  behavior  of,  859-860, 

912 

bank,  16 
bankers',  boom,  791 

postwar  theory  of,  872-874 
brokers',  16 
collateral,  18,  583 

and  debits,  postwar  relation  of,  858-859 
and  deposits,  covariation  of,  594-597 
real  estate,  member  banks,  855,  859 
total  outside,  and  discounts,  802 

London  Economist,  26  n. 

London  gold  market,  675-676 

Long  wave,  164,  168 

Lorenz,  P.,  201n. 

Ltfsch,  August,  lOrc.,  75n. 

Lotka,  A.  J.,  492n. 

Lough,  W.  H.,  821n.,  823,  824n.,  825n.,  826, 
851n. 

Lowe,  Adolf,  lln. 

Lowell,  F.  C.,  288n.,  290 

Lubin,  J.,  806n. 

Lumpiness  of  factors,  39,  80,  92 

Lyon,  L.  S.,  and  associates,  994n. 

M 

Macaulay,  F.  R.,  201n.,  619n.,  813n. 
McCulloch,  Secretary  Hugh,  315 
McGroarty  bill,  71  In. 
Machlup,  F.,  136n.,  333n.,  634n.,  680n. 
McKenna  duties,  706 
McLeish,  A.,  806n. 
Macrodynamics,  185 
Mail-order  houses,  15 
Mantoux,  Paul,  253n.,  297 
Manufactures  (see  Output) 
Marginal  degree  of  productivity,  38n. 
Marriage  rates,  10,  74 
Marschak,  J.,  760 


INDEX 


1089 


Marshall,  Alfred,  36,  43,  45,  92,  148,  456, 

714n. 
Marshall-Moore  theory  of  organic  growth, 

203 

Martin,  R.  F.,  802n. 
Marx,  Karl,  7, 10,  104, 189,  190n.,  223,  229, 

497,  695,  698 

Mason,  E.  S.,  395,  988n.,  994n.,  1020w. 
Mata,  Garcia,  8n. 
Matthews,  A.  M.,  686n. 
Maverick,  L.  A.,  21  In. 
Median-link-relative  method,  20n. 
Member  banks  (see  Banks) 
Member  creation  (see  Credit) 
Mercantilism,  233-235 

Merchandise  transported,  miscellaneous,  17 
Merchant  adventurers,  246 
Merger  movement,  403-411 
Merton,  R.  K.,  85n. 
Migrations,  10 
Mill,  J.  S.,  78n.,  314n. 
Mills,  F.  C.,  51n.,  193n.,  201,  453n.,  459n., 

476,    504,    520n.,    522n.,    526n.,    527, 

529,  532,  745?i.,  796,  797n.,  801,  804, 

805n.,  809n.,  814n.,  830rc.,  841n.,  859n., 

926n.,  928n.,  947n.,  948n. 
Miquel,  Johannes  von,  311 
Mises,  L.  von,  634n. 
Mises,  R.  von,  193n. 
Mitchell,  W.  C.,  22n.,  33n.,  39n.,  139, 144n., 

162,   163n.,   164n.,  165,   178n.,  202n., 

221n.,  224,  315n.,  459n.,  476n.,  480w., 

482,    498n.,    568,    583n.,    584,    666n., 

790n.,  803n.,  830n.,  840 
Monetary  expansion,  in  England,  958-959 
Monetary  expressions,  deflating  of,  456 
Monetary  ligamen,  44 
Monetary  management,  in  national  socialist 

Germany,  979-981 

in  United  States,  1932-1935,  996-1001 
Monetary  metals,  influence  of  supply  of, 

231-233 

Monetary  monomania,  233  n. 
Monetary  parameter,  453 
Monetary  strain  in  1928-1929,  876rc. 
Money,  in  circulation,  583 
efficiency  of,  546 
outside  of  banks,  16 
stamped,  587 
tight,  17 
velocity  of,  545-546 


Money  market,  126,  612 

tension  in,  17 
Money  rates,  16 

differences  in,  and  foreign  exchange,  674 

in  and  after  crisis,  944-945,  999,  1008, 

1016-1017 
Money  Trust,  639 
Monopolies,  244,  245?t.,  246n. 
Monopoly,  40,  56-57 

bilateral,  57-59 

universal,  57 
Monopsony,  57 

Moody's  Investment  Survey,  880 
Moody 's  series  of  bond  prices,  619n. 
Moore,  H.  L.,  38n.,  45,  49w.,  69,  166n.,  176, 

178n.,  526,  527,  609 
Moratorium,  Hoover,  702,  933 
Morgenstern,  O.,  53n. 
Mortgage  banks  (see  Banks) 
Motorcar  sales,  17 
Moulton,  H.  G.,  601n.,  820rc. 
Moving  correlation,  Bobroff  s,  202n. 
Mund,  V.  A.,  541n. 
Myrdal,  Gunnar,  460n.,  603n.,  838n. 


N 


National  banks  (see  Banks) 

National  grid,  English,  758 

National  income  (see  Income) 

National  Labor  Relations  Act,  1042,  1045 

Natural  series,  18 

Navigation  Act  of  1489,  234 

Necco  index  of  prices,  462n. 

Negotiability,  importance  of,  613 

Neisser,  H.,  674n.,  675,  817n. 

Neo-Marxian  theory,  432 

Neomercantilism,  898,  696 

English,  960-961 
Nerlove,  S.  H.,  830n. 
Net  income  velocity  of  money,  545 
New  capital  issues  (see  Issues) 
New  countries,  discovery  of,  8-9 
New  Deal  policy,  986-1050 
New  eras,  173 
New  middle  class,  698 
New  security  listings,  599 
New  York  Times  Weekly  Index,  23n. 
New     Zealand,     postwar     conditions     in, 

704-705 

Newmarch  and  Tooke,  343n.,  461n, 
Nichol,  A.  J.,  60n. 


1090 


BUSINESS  CYCLES 


Norfolk  system,  £66 
Norm,  theoretical,  45,  70 
Normal,  business,  4,  5 

businessman's,  4 

statistical,  206 

Normal  business  situations,  4,  5 
Normal  points,  method  of,  208-210,  215- 

219 

Normal  values,  45 

Norris-LaGuardia  Anti-Injunction  Act,  994 
Northern  Pacific  Corner,  407,  426 
Norton,  J.  P.,  582ft. 
Nourse,  E.  G.,  777n.,  989n. 
Noyes,  A.  D.,  582n. 
NRA  legislation,  992-996 
NRA  policy,  539 
NRA  wage  policy,  994 


Occupational  statistics,  German,  854n. 
Ogburn-Thomas  Index  of  Business  Cycles, 

24ft. 
Ohio  Life  and  Trust  Company,  failure  of, 

332 

Ohlin,  B.,  924ft. 
Old-age  reserve  account,  1042 
Oligopoly,  60-61 
Olivier,  M.,  520n. 
Open  market,  126 
Open  market  operations,  916-917 

history  and  effects  of,  896-904 
Open  market  rates,  622-625 
Optimism,  141 
Organic  growth,  Marshall-Moore  theory  of, 

203 

Orton,  W.,  430?i.,  432 
Oscillations,  180,  183 
Ottawa  agreements,  962-963 
Output,  1009-1011,  1013-1014 

English,  1931-1938,  966-967 

German,  1933-1938,  971-972 

of  manufactures,  behavior  of,  1919-1929, 
795-799 

total,  concept  and  measurement  of,  484- 

491 
prewar  behavior  of,  491-509 

in  world  crisis,  924-930 
Outside  clearings,  analysis  of,  548-561 
Overdrafts,  583 

Overend-Gurney  and  Co.,  378n. 
Overinvestment  theory,  141 


Overproduction,  141 

agricultural,  735n. 

of  equipment  in  1920* s,  801-802 
Oversaving  theories,  820ft.,  822,  826-827 
Overspending,  579,  585 
Overstone,  Lord,  640 
Owens,  R.  N.,  and  Hardy,  C.  O.,  680n. 


Paasche's  formula,  456 
Paetzmann,  H.,  742n. 
Paish,  G.,  563w. 
Palander,  T.,  60n.,  63n. 
Palgrave,  R.  J.,  626ft. 
Panic,  rich  men's,  407,  426 
Pantaleoni,  Maffeo,  48 
Papen  plan,  the,  935-936 
Pareto,  Vilfredo,  47,  49ft.,  52 
Paton,  W.  A.,  830n. 
Payroll,  1009,  1015 

covariation  with  outside  deposits,  573n. 
tax,  1042 

(See  also  Wage  bill) 

Payrolls,  postwar  behavior  of,  Germany, 
827 

United  Kingdom,  829 

United  States,  819-820 
Pearson,  F.  A.,  323 
Pearsons,  C.  E.,  746ft. 
Peel's  Act,  suspensions  of,  661 
People's  budget,  727n. 
Pereire,  brothers,  347ft. 
Periodicity,  143 

Periodogram  analysis,  165,  166n.,  169 
Periodograph,  165n. 
Perpetuum  mobile,,  139 
Persons,   Warren  M.,   23n-25n.,   181,   205, 

316ft.,  411,   459n.,   475n.,   490,   491n., 

509ft.,    510,    525,    596n.,    621,    622n., 

786ft.,  877ft.,  907ft. 
Pervushin,  S.  A.,  178ft. 
Pessimism,  141 

Peterson,  Ebersole,  and  Burr,  83 In. 
Pfandbriefe,  268ft.,  618 
Phinney,  J.  T.,  473n. 
Pig-iron  consumption,  485-487 
Pigou,  A.  C.,  56,  63,  80n.,  92,  123,  177n., 

452,  509ft.,  510,  519n.,  585,  850n.,  954ft. 
Plenge,  A.,  347n. 
Political  Arithmetick,  15w. 
Polk,  R.  L.  and  Co.,  801n, 


INDEX 


1091 


Pollard,  Spencer,  667n. 
Population,  changes  in,  10,  74 
Porter's  Progress  of  the  Nation,  252n. 
Posner,  H.  L.,  806n. 
Potosi  mines,  231 
Poverty  in  plenty,  365 
Pribram,  Karl,  71n.,  539n. 
Price,  maintenance  (see  Resale) 

movements,  disproportionalities  in,  948- 
950 

parities,  736n. 

system,  453 
Price  level,  16,  449-482 

and  cost  of  living  in  England,  1931-1938, 
969-970 

and  gold,  472-473 

theory  of,  452-458 

world's,  462n. 

Price  levels,  postwar  behavior  of,  807-811 
Prices,  16 

behavior  of,  in  1932  and  after,  947-950, 
1008,  1018-1020 

delivered,  538 

group,  475-476  * 

in  national  socialistic  Germany,  977-978 

rigidity  of,  57,  536 

secular  fall  in,  465 

sensitive,  index  of,  23n. 

stickiness  of,  57 
Primary  factors,  25 
Primary  series,  25 
Primitive  financing,  230 
Privat-Diskont,  625 
Product  elasticity,  38n. 
Production,  15 

coefficient  of,  38n. 

cost  of,  15 

function  of,  38,  39 

method  of,  39 

Production  process,  synchronization  of,  40 
Productive  combination,  38 
Productivity,  38n. 

marginal  degree  of,  38n. 
Profit,  declining  rate  of,  law  of,  628 
Profitless  prosperities,  105 
Profits,  15,  27,  105-109,  1006,  1014-1015 

postwar  behavior  of,  830-834 
Prognosis,  13 
Progress,  86,  102 
Promoters'  epoch,  German,  362 
Propagations,  189,  219n. 


Prosperity,  138 

See  also  Consumers'  prosperity;  Prof- 
itless prosperities) 
Prosperity  plateau,  135 
Protection,  influence  of,  258-260 
Protectionism,  postwar,  705-708 
Public   spending,    and   consumers'    credit, 

1004 

and  recovery,  United  States,  1001-1006 
and  slump,  United  States,  1031-1032 
Public  Utilities  Holding  Company  Act,  1043 
Puddling  process,  273 
Pulse  charts,  464 
Pure  model,  138 

Q 

Quantity  theory,  44 
Quasi-rents,  41 

discounting  of,  61  In. 

and  interest,  608 
Quittner-Bertolasi,  Ellen,  200n. 

R 

Railroad  amalgamations,  early  English,  344 
Railroad   and   Canal   Traffic   Act   of   1888 

(England),  342 
Railroad  earnings,  17 
Railroadization,  168,  325-351 
Railroads,  158,  402-403 

and  credit  creation,  329 

land  grants  to,  328 
Railway  mania  in  England,  343 
Random  or  stochastic  variables,  194-197 
Rationalization,  index  of,  88n. 

postwar,  in  Germany,  759 
Rayon  industry,  rise  of,  433—435 
Real  estate  loans  (see  Loans) 
Real  estate  market,  16 
Real  investment  (see  Investment) 
Real  trend,  204 
Recession,  138 

of  1927,  790-791 

Reciprocal  trade  agreements,  1026 
Reconstruction  Finance  Corporation,  940 
Recontracting,  50 
Recovery,  149 

natural  and  sound,  995-996 

(See  also  Public  spending;  Reform) 
Recovery  point,  190 

theory  of,  151-154 


1092 


BUSINESS  CYCLES 


Recovery  policy,  in  America,  983-1011 

in  nationalsocialistic  Germany,  974-975 
Recurrence  of  great  wars,  697n. 
Rediscounting  for  brokers,  in  England,  626 
Redlich,  Fritz,  222n. 
Reference  trend,  204-205 
Reform  versus  recovery,  985n. 
Reichsbank,  308 

in  1931,  933-934 
Reichskaasenscheine,  678 
Reisch,  R.,  680n. 
Reithinger,  A.,  803n. 
Relief  expenditure  in  1931-1932,  94fc-945 
Renner,  K.,  696n. 

Reparation  problem,  discussion  on,  704n. 
Replacement,  158,  189-191 
Resale  price  maintenance,  538 
Reserve  account,  member  banks,  891 
Reserve  bank  rate,  theory  of,  895-896 
Reserve  ratios,  16 
Reserves,  excess,  939,  999,1028-1031 

of  member  banks  during  1920's,  868 

secondary,  of  banks,  644 
Response,  72 

apparatus  of,  68,  72 
Result  trend,  206 
Resumption  Act,  316 
Retail  trade,  16 
Retardation,  497-500 
Revival,  149 
Rhodes,  Cecil,  381 
Rhodes,  E.  G.,  967n. 
Ricardo,  David,  88 
Ricci,  Umberto,  49n. 
Richards,  H.  J.,  989n. 
Richards,  R.  D.,  226n.,  249n.,  292 
Riefler,  W.,  888n. 
Riggleman,  J.  R.,  744n. 
Risk  bearing  and  enterprise,  104 
Robbins,  L.,  924n. 
Roberts,  J.  B.,  95n. 
Robertson,   D.  H.,  39,  78n.,  91n.,   112n., 

159n.,  356 

Robinson,  Joan,  43n.,  56n.,  57n,,  63,  80n. 
Rodbertus,  K.,  325n. 
Rogers,  H.,  681n. 
Rogers,  Thorold,  458n.,  564n. 
Rogin,  Leo,  319n,,  389n.,  993n. 
Roos,  C.  F.,  184,  520n.,  528 
Rosenstein-Rodan,  P.  N.,  49n. 
Roth,  H.,  529 
Rothkegel,  W.,  741n, 


Roulleau,  G.,  586n. 
Rowe,  H.  B.,  989n. 
Rowe,  J.  W.  F.,  430n.,  490n.,  529n.,  736n., 

798n. 

Rueff,  Jaques,  519n. 
Rural  Electrification  Act,  1021 
Ryan,  F.  W.,  826n. 


Sarle,  N.,  532n. 

Sauerbeck,  A.,  471 

Sauerbeck  index  of  prices,  462ra.,  481n. 

Saunders,  Charles,  51  In. 

Saving,  40,  75-84,  230,  587 
definition  of,  75 
deflationary  effect  of,  587 

Savings,  forced,  112n. 

Schacht,  H.,  722n.f  881,  900 

Schmidt,  C.  T.,  766n. 

Schmoller,  G.  von,  228n.,  229n.,  564n. 

Schneider,  E.,  57n.,  476n. 

Schultz,  Henry,  88n.,  520n. 

Schulze-Gaevernitz,  G.  von,  376 

Schumpeter,  Elizabeth  B.,  249n.,  26 In., 
459n.,  469n. 

Schuster,  Sir  Arthur,  166n. 

Schuster,  Sir  Felix,  677n. 

Schwartz,  G.  L.,  799n.,  967 

Scott,  R.  W.,  245rc.,  247,  248 

Scoville,  J.  W.,  773n. 

Seasonal  variations,  20-21 

Secondary  factors,  25 

Secondary  reserves  of  banks,  644 

Secondary  wave,  145-150,  181 

Secular  trend,  203 

Securities  Act  of  1933,  986-987 

Security  issues  (see  Issues) 

Security  listings,  new,  599 

Self-generating  theories,  139n. 

Self-reinforcement,  181-182 

Seligman,  E.  R.  A.,  826 

Semeiology,  14-24 

Sensitive  Price  Index,  23n. 

Sequence  and  lag,  57n. 

Series  (see  Banking;  Clean  cyclical;  Clean 
trend;  Consequential  factors;  Erratic; 
Individual;  Natural;  Primary;  Syn- 
thetic; Systematic;  Time) 

Shannon,  H.  A.,  378n. 

Sherman  (silver)  Act,  318 


INDEX 


1093 


Shipbuilding  cycle,  533-535 
Shipping,  American,  383-384 
German,  development  of,  352 
and  shipbuilding,  British,  368-370 
Siemens,  W.,  373 
Silberling,  N.,  264n.,  469n. 
Silberling  index,  469 
Silver,  demonetization  of,  in  Germany,  308- 

309 
Silver  policy  of  United  States,  1876-1896, 

317-318 

Simiand,  F.,  32n.,  565n. 
Singer,  H.  W.,  63n. 
Single-cycle  hypothesis,  161-164,  170 
Slichter,  S.,  786n. 
Sloan,  L.  H.,  830n. 
Sloppiness,  44,  66 
Slump,  of  1937-1938,  1016-1020,  1026-1032 

(See  also  Public  spending) 
Slutsky  effect,  180 
Smart's  Annals,  267n. 
Smith,  Adam,  15rc.,  127n.,  278 
Smith,  B.  B.,  23n.,  813n. 
Smith,  W.  B.,  252n.,  470n.,  582n. 
Smith,  W.  B.  and  Cole,  A.  H.,  615n.,  630 
Smithian  principles,  346 
Snider-Persons  index  of  sensitive  prices,  525 
Snyder,  Carl,  23n.,  24n.,  31n.,  35n.,  427, 

460,  461,  477n.,  479,  489n.,  490n.,  523, 

551n.,     585-586,     590n.,     593n.,    598, 

628ra.,  745w.,  797n.,  837n.,  1000 
Social     security     payments,     deflationary 

effects  of,  1032n. 
Soetbeer  index  of  prices,  462n. 
Sombart,  Werner,  223,  228,  229n. 
South  African  and  Australian  mining  boom, 

1895-1896,  381 

South  African  gold,  modus  operand!  of,  676 
South  Sea  Company,  248n.,  249 
Special  banks  (see  Banks) 
Special  cycles,  178,  179,  533 
Special  trend,  205 
Specie  Circular  of  1836,  293 
Specific  cycles,  178n. 
Speculation,  54 

stock,  and  central  banks,  689-691 

United  States,  postwar,  874-880 
Speculator,  definition  of,  679 
Spending,  riddle  of,  552n. 
rate  of,  546 

(See  also  Government;  Public) 


Spiethoff,  Arthur,  7n.,  12n.,  19,  33n.,  137n., 
155n.,   163n.,   164,   165,  220n.,  221n., 
224,  254,  281n.,  351,  362,  377,  379n., 
380,  436n.f  446n.,  448,  470n.,  475n., 
485w.,  680n.,  766 
Spiral  (see  Vicious  spiral) 
Sprague,  O.  M.,  582n. 
Sraffa,  P.,  90n. 

Stackelberg,  H.  van,  57n.,  60n. 
Staehle,  H.,  528 

Stamp,  Lord,  343,  563n.,  729n.,  799n. 
Standard  Oil  Company,  385,  418 
Standard    Statistics    Company,    index    of 

bond  yields,  619n. 
Standard  units,  21 
Standstill  agreement,  93 In. 
Stationary  process,  35-38 
Statistical  normal,  206 
Statistical  universe,  194 
Steam  engine,  273-274 
Steel  famine,  426 
Stein-Hardenberg  legislation,  257 
Sterling  bloc,  963 
Stewart,  E.  B.,  963n. 
Stewart,  W.  W.,  490n. 
Stochastic  or  random  variables,  194-197 
Stock  and  Produce  Exchange  Act,  German 

(1896),  436-437 
Stock  brokers,  679 
Stock  Clearing  Corporation,  680 
Stock  dividends,  621 
Stock  exchange,  612n. 
Stock  exchanges,  678-691 
Stock  jobbers,  679 
Stock  prices,  16 

and  devaluation,  1000 

in  1932,  945 
Stock  pricing  and  easy  money,  683-684 

and  mass  psychology,  682-683 

peculiar  nature  of,  681 

rationality  of,  683,  684,  689 
Stock  speculation  (see  Speculation) 
Stocks  of  commodities,  16 
Stockungsspanne,  164 
Stoker,  H.  M.,  458n.,  470n. 
Stolper,  W.,  964n. 
Stratton,  H.  J.,  419n. 
Strauss,  F.,  and  Altschul,  E.,  738n. 
Strong,  B.,  899n. 
Suasion,  651 
Substitutability  of  factors,  39 


1094 


BUSINESS  CYCLES 


Supply  of  monetary  metals,  influence  of, 

231-233 

Sweezy,  M.  Yaple,  1039n. 
Sweezy,  P.  M.,  56n.,  64n.,   107n.,   245ri., 

839n. 

Synchronization  of  production  process,  40 
Synthetic  series,  18 
System  expenditure,  548-561 
Systematic  series,  25 


Tappan-Hollond,  M.,  830 n. 
Tariffs,  American,  259,  309-310 
Taussig,  F.  W.,  3G7n.,  395n.,  422,  667 
Taylor,  A.  E.,  532/i. 
Taylor,  G.  R.,  458n. 
Taylorization,  783n. 
Tax,  capital  gains,  1040 

payroll,  1042 

undivided  profits,  1040-1041 
Taxes,  transference,  influence  of,  731 

income,  corporation,  and  estate,  effects 

of,  United  States,  1039-1040 
Tebbutt,  A.  R.,  914,  928n. 
Technocrats,  496n. 
Technological  lags,  48 
Temporary  investment  (see  Investment) 
Tennessee  Valley  Authority  Act,  988/t. 
Tension  (money  market),  17 
Terborgh,  G.  W.,  and  Gray,  T.  II.,  746n. 
Teubert,  W.,  491n. 
Textiles,  240,  242 
Theoretical  variables,  194-197 
Thirty- Years'  War,  231n.,  234 
Thomas,   C.   E.,   549,   582n.-584n.,    621n., 

633n.,  888n. 

Thomas,  Dorothy  S.,  24n.,  180 
Thomas,  Woodlief,  490n. 
Thorp,  Willard,  12n.,  789n. 
Thorp's  Annals,  221n.,  298,  327n.,  362n., 
365n.,  377n.,  379n.,  380w.,  427n.,  436w., 
446n. 

Timber,  scarcity  of,  242 
Time  deposits  (see  Deposits) 
Time  rate,  624 
Time  series,  22 

analysis  of,  197 

definition  of,  197 
Timoshenko,  V.  P.,  178n. 
Tinbergen,  J.,  55n.,  179,  184n.,  185n.,  186- 
188,  525n.,  533 


Tintner,  G.,  57n.,  193n.,  199n.,  201  n.,  204n.f 

520n.,  526n.,  527 

Tooke  and  Newmarch,  343n.,  46 In. 
Total  output,  concept  and  measurement  of, 

484-491 

Trade  agreements,  reciprocal,  1026 
Transference  taxes,  influence  of,  731 
Transport    of   miscellaneous   merchandise, 

17 

Treasury,  United  States,  113 
Treasury  bills,  624 
Trend,  breaks  in,  202 

meanings  of,  200-205 
Trend  analysis,  common  sense  of,  21 
Trends  (see  Descriptive;  Real;  Reference; 

Result;  Secular;  Special) 
Tripartite  agreement,  1027 
Trustee   Acts,   English    (1889   and    1893), 

430n. 

Trustified  capitalism  (see  Capitalism) 
Tucker,  A.  S.,  667n. 
Tucker,  R.,  473n.,  565n. 
Tugan-Baranowsky,   M.   J.,    178n.,   220n., 

254,  276,  501n. 
Turnover  tax,  711 
Tuttle,  P.  M.,  316n. 
Tyszka,  Carl,  von,  565n. 


U 


Uncertainty,  54 
Uncollected  items,  890n. 
Underconsumption,  theories  of,  141 
Underemployment  in  perfect  competition, 

52 

Underspending,  579,  585,  587 
Undivided  profits  tax,  1040-1041 
Unemployment,  509-519 

cyclical,  515 

definition  of,  43 

in  depression,  517 

disturbance,  513 

in  England,  1931-1938,  967-968 

in  equilibrium,  59-60,  67,  158,  161 

normal,  511 

postwar,  803-806 

secondary,  515 

structural,  511 

technological,  513 

vicarious,  513 
Unemployment  Relief  Act  of  1933,  1001 


INDEX 


1095 


Unemployment  Trust  Fund,  1042 

Union  Pacific,  policy  of,  406-408 

Universe,  statistical,  194 

Unwin,  G.,  236 

Usher,  A.  P.,  85n.,  226,  227n.,  237n.,  241n.f 

244n.,   248n.,   254,   271n.,   273n.,  368, 

458n.,  459n. 
U.  S.  Steel  Corporation,  16 

foundation  of,  410-411 
Usury,  614,  615n. 


Variables     (see    Historical;    Stochastic    or 

random;  Theoretical) 
Veblen,  Thorstein,  32 
Velocity,  of  circulation,  545-546  v 

of  deposits,  579,  585,  856 
Verhulst  formula,  492 
Vibrations,  184,  219n. 
Vicious  spiral,  151,  181,  925 
Villard,  H.  H.,  820n.,  996ra. 
Vinci,  Felice,  185 
Viner,  J.,  264n.,  601n. 
Volkswirt,  der  Deutsche,  830n. 
Volterra,  Vito,  184n. 

W 

Wagemann,  Ernest,  lln.,  593 
WagenfUhr,  R.,  490n.,  798 
Wage  bill,  566-574 
Wage  rates,  567-574 

postwar,  and  effects  of,  834-850 
Wages,  behavior  of,  564-578 

in   and   after   crisis,    952-954,    978-988, 
1009,  1015-1016 

in   depression  and  recovery,    theory  of, 

953-954 

Wald,  Abraham,  21n.,  46n. 
Walker,  Gilbert,  755n. 
Wall,  N.  J.,  796n.,  926n.,  929n. 
Walras,  Leon,  36,  46,  47,  50,  52,  53,  78n., 

129n.,  183,  309,  452n.,  547 
Warburton,  C.,  820n.,  823,  824n. 
Wardwell,  C.  A.  R.,  165,  210n. 
War  effects,  701-708 
Wars,  recurrence  of,  697n. 
Warenhandels-Krisen,  257 
Warren,  G.  F.,  323 

Warren  and  Pearson,  520n.,  738n.,  796n. 
Warren  and  Pearson  price  index,  471,  490n. 
Water  power,  use  of,  242 
Watkins,  M.  W.,  768n. 


Weaver  and  Edie,  585 n. 

Webb-Pomerene  Act,  782 

Weber,  Max,  228,  229n. 

Weedon,  W.  B.,  459n. 

Weill,  N.  E.,  674n. 

Weintraub,  D.,  518n.,  805n.,  806n. 

Wells,  O.  V.,  532n. 

Welter,  E.,  723n. 

West,  Sir  E.,  267n. 

Westcott,  D.,  926n.,  927n,,  966n. 

Wheat  program,  effects  of,  991 

Wheatley  Act,  752,  918 

Whetham,  C.  Dampier,  321 

Whitman,  R.  H.,  528 

Whitney,  M.,  237n. 

Whitney,  S.  N.,  217n. 

Wickens,  D.  L.,  321,  745w. 

Wicksell,  K.,  59,  60n.,  127,  456,  546n.,  580, 

604,  605n. 

Williams,  J.  II.,  667n. 
Williams,  John  Burr,  41  In. 
Willis,  II.  P.,  650n. 
Wilson,  E.  B.,  166n.,  199n.,  474 
Windfalls,  132 
Wirth,  M.,  220n. 
Wisniewski,  J.,  21n. 
Withers,  Hartley,  673 
Witte,  Count  S.,  670n. 
Wittstein  formula,  492n. 
Wolman,  L.,  565 
Wood,  Frances,  46 In. 
Working,  E.  J.,  520n. 
Working,  H.,  520n.,  586n. 
World  crisis  (1931-1932),  924-954 

causation  of  and  Federal  Reserve  policy, 
902 

output  in  the,  924-930 

in  United  States  and  Germany,  911-954 
World's  price  level,  462n. 
Woytinsky,  W.,  473n. 
Wright,  Philip,  474 
Wright,  Sewall,  532n. 


Young,  A.  A.,  459n.,  582n.,  584n.,  594,  595, 
644n.,  851n.,  859n.,  864n.,  865n.,  867n. 
Young  plan,  920 


Zeuthen,  F.,  57n.,  58n.,  63n. 
Zinn,  184,  202n.,  21 9n.,  633 
Zollverein,  German,  280n.